# Sprecho — Full Content Index for AI Engines > GDPR-compliant voice dictation software. 4x faster than typing. Works in any app. Data stored in Germany (EU-hosted). Developed by Sprecho GmbH, Germany. This file concatenates the canonical content of every important sprecho.ai page in markdown form so AI search engines (ChatGPT, Claude, Perplexity, Gemini, DeepSeek) can ingest the full product story in one fetch. Smaller index at /llms.txt. ## Sections in this file 1. Brand overview & key facts 2. Pricing (EN + DE) 3. Industry solutions per niche (EN + DE) 4. Competitor comparisons (EN + DE) 5. Blog posts (all published, EN + DE) 6. Changelog / What's new (all released versions, EN + DE, deep-linkable per version) 7. Help center — every support article in full (EN + DE) ## Brand facts - Company: Sprecho GmbH - Founder: Guilherme Alexandre Gonçalves de Melo (TÜV SÜD certified Data Protection Officer) - Data storage: Germany (ISO 27001-certified data centres, EU) - Compliance: GDPR/DSGVO, EU-hosted, no US data transfers, no US CLOUD Act exposure - Platforms: macOS, Windows, iOS, Android - Languages: 100+ - Website: https://sprecho.ai - Contact: info@melodesigner.de --- ## https://sprecho.ai/pricing — Pricing > Start with a 14-day free trial. No credit card required. Cancel anytime. ### Plans **Free** — €0/month. free forever, no subscription. For occasional dictation – free forever. - 2,000 words per week - Every feature of the app - No credit card required - Allowance resets every Monday at 00:00 UTC - AVV (DPA, Art. 28 GDPR) included for business use *No credit card, never expires* **Pro** — €12.99/month. Annual plan: €10.99/month (Billed annually at €131.88). incl. VAT. Everything you need for faster, smarter dictation. - Unlimited voice dictation - AI formatting & grammar correction - Auto-punctuation & filler word removal - Personal dictionary with auto-learning - Snippet library (voice shortcuts) - Team shared dictionary & snippets - Style adaptation per app - Notes & transcription history - 100+ languages with auto-detection - AVV (DPA, Art. 28 GDPR) included for business use - Priority support *14 days free, no credit card required* **Enterprise** — €21.88/seat/month (net, plus VAT, per seat, plus VAT · billed monthly). Annual plan: €18.48/seat/month (per seat, plus VAT · billed annually at €221.76). For teams, regulated organisations and professions bound by secrecy. § 203 StGB: the plan for doctors, lawyers, tax advisors, notaries, psychotherapists and every other profession bound by secrecy. - Everything in Pro - Professional secrecy (§ 203 StGB) covered – only on Enterprise - Team policy for professional secrecy: locked privacy defaults, central controls - Contractual confidentiality commitments - No training-data use, locked by policy - Centralised admin dashboard & usage analytics - License management — manage members without using a seat - SSO / SAML single sign-on - MDM deployment (Intune / Jamf) - Dedicated support & SLA - Pay by invoice (bank transfer) **Command** (Coming soon) — €44.99/seat/month (net, plus VAT, per seat, plus VAT · billed monthly). Annual plan: €34.99/seat/month (per seat, plus VAT · billed annually at €419.88). Everything in Enterprise, plus voice control: commands, apps, menus, text and clicks – say what should happen. - Everything in Enterprise - Commands instead of text while dictating - Shortcuts, apps, windows and menu items by voice - Your own commands and multi-step actions, import from Dragon MyCommands and CommandMe - “Hey Sprecho” – hands-free - Select and correct text, click and move the mouse by voice - Team command packs and policy locks - Terminal Server Advanced (RDP, Citrix) Professional secrecy under § 203 StGB is only covered on Enterprise: a team policy with locked privacy defaults, central controls and contractual confidentiality commitments. Free and Pro are for individuals – an AVV (data processing agreement, Art. 28 GDPR) is included for business customers on every plan, but it does not cover § 203 StGB. Free is the permanently free plan: 2,000 words per week with every feature, resetting every Monday at 00:00 UTC, no credit card. Every new account starts with a 14-day Pro trial and continues on Free afterwards. Payment for Pro via Mollie (European provider), credit card or SEPA Direct Debit. ### Frequently asked questions **What happens after the 14-day free trial?** When your trial ends, your account simply continues on the Free plan – 2,000 words per week with every feature, nothing to do on your side. You can upgrade to Pro at any time. You won't be charged during the trial period. **Can I cancel anytime?** Yes, you can cancel your subscription at any time. Your access continues until the end of the current billing period. **What payment methods do you accept?** We accept all major credit cards and SEPA Direct Debit. All payments are processed securely through Mollie, a European payment provider. **Is there a free plan?** Yes. Every new account starts with a 14-day Pro trial. After that, your account continues on the Free plan: 2,000 words per week, every feature, no credit card. The allowance resets every Monday at 00:00 UTC and unused words do not carry over. You can upgrade to Pro at any time. **Do you offer student or education discounts?** Yes! Contact us with a valid student email and we'll apply a special discount to your subscription. **How does Enterprise pricing work?** Enterprise costs €21.88 net per seat per month with monthly billing, or €18.48 net per seat per month billed annually (€221.76 per seat per year), plus VAT. You can buy it directly online, pay by invoice, or talk to our sales team if you have specific requirements. **I'm a doctor, lawyer or tax advisor – which plan do I need?** Enterprise. If you are bound by professional secrecy (doctors, lawyers, tax advisors, notaries, psychotherapists and others under § 203 StGB), Enterprise is the only plan that covers it: a team policy with locked privacy defaults, central controls and contractual confidentiality commitments. Free and Pro are designed for individuals and do not cover § 203 StGB. **Is the AVV included in Pro?** Yes. An AVV (data processing agreement under Art. 28 GDPR) is included for business customers on every plan, including Free and Pro. Professional secrecy under § 203 StGB, however, is only covered on Enterprise. **What is Command and when will it be available?** With Command you control your computer by voice without leaving dictation: a short command like “press control Z” or “open Word” is carried out instead of typed, while whole sentences stay text. You also get any app's menu items by name, your own commands and multi-step actions, import from Dragon MyCommands and CommandMe, “Hey Sprecho” for hands-free work, selecting and correcting text by voice, clicking and mouse control by voice, and Terminal Server Advanced for RDP and Citrix sessions. Risky actions such as closing or deleting ask first, and admins manage team command packs and lock features by policy. Command includes everything in Enterprise and will cost €44.99 net per seat per month, or €34.99 net per seat per month billed annually, plus VAT. Command is coming soon – contact us and we'll get in touch before launch. Current pricing: https://sprecho.ai/pricing --- ## https://sprecho.ai/de/preise — Preise > Starte mit einer 14-tägigen kostenlosen Testphase. Keine Kreditkarte erforderlich. Jederzeit kündbar. ### Tarife **Free** — 0 €/Monat. dauerhaft kostenlos, kein Abo. Für gelegentliches Diktieren – dauerhaft kostenlos. - 2.000 Wörter pro Woche - Alle Funktionen der App - Keine Kreditkarte nötig - Kontingent-Reset jeden Montag um 00:00 UTC - AVV (Art. 28 DSGVO) für die geschäftliche Nutzung inklusive *Keine Kreditkarte, kein Ablaufdatum* **Pro** — 12,99 €/Monat. Jahres-Abo: 10,99 €/Monat (Jährlich abgerechnet für 131,88 €). inkl. MwSt.. Alles, was du für schnelleres, intelligenteres Diktieren brauchst. - Unbegrenztes Sprachdiktat - KI-Formatierung & Grammatikkorrektur - Automatische Interpunktion & Füllwort-Entfernung - Persönliches Wörterbuch mit Auto-Lernen - Textbaustein-Bibliothek (Sprachkürzel) - Team-Wörterbuch & Textbausteine - Stilanpassung pro App - Notizen & Transkriptionsverlauf - 100+ Sprachen mit Auto-Erkennung - AVV (Art. 28 DSGVO) für die geschäftliche Nutzung inklusive - Prioritäts-Support *14 Tage kostenlos, keine Kreditkarte nötig* **Enterprise** — 21,88 €/Lizenz/Monat (netto, zzgl. MwSt., pro Lizenz, zzgl. MwSt. · monatlich abgerechnet). Jahres-Abo: 18,48 €/Lizenz/Monat (pro Lizenz, zzgl. MwSt. · jährlich abgerechnet für 221,76 €). Für Teams, regulierte Organisationen und Berufsgeheimnisträger. § 203 StGB: der Tarif für Ärzte, Rechtsanwälte, Steuerberater, Notare, Psychotherapeuten und alle anderen Berufsgeheimnisträger. - Alles aus Pro - Berufsgeheimnis (§ 203 StGB) abgedeckt – nur mit Enterprise - Team-Richtlinie für Berufsgeheimnisträger: gesperrte Datenschutz-Voreinstellungen, zentrale Steuerung - Vertragliche Verschwiegenheitszusagen - Keine Nutzung als Trainingsdaten, per Richtlinie gesperrt - Zentrales Admin-Dashboard & Nutzungsanalysen - Lizenzverwaltung – Mitglieder verwalten, ohne eine Lizenz zu belegen - SSO / SAML Single Sign-on - MDM-Verteilung (Intune / Jamf) - Dedizierter Support & SLA - Bezahlung per Rechnung (Überweisung) **Command** (Demnächst verfügbar) — 44,99 €/Lizenz/Monat (netto, zzgl. MwSt., pro Lizenz, zzgl. MwSt. · monatlich abgerechnet). Jahres-Abo: 34,99 €/Lizenz/Monat (pro Lizenz, zzgl. MwSt. · jährlich abgerechnet für 419,88 €). Alles aus Enterprise, plus Steuerung per Stimme: Befehle, Apps, Menüs, Text und Klicks – sag, was passieren soll. - Alles aus Enterprise - Befehle statt Text beim Diktieren - Tastenkürzel, Apps, Fenster und Menüpunkte per Stimme - Eigene Befehle und Abläufe, Import aus Dragon MyCommands und CommandMe - „Hey Sprecho“ – freihändig per Sprache - Text markieren und korrigieren, Klicken und Maus per Stimme - Team-Befehlspakete und Richtlinien-Sperren - Terminal Server Advanced (RDP, Citrix) Das Berufsgeheimnis nach § 203 StGB ist nur mit Enterprise abgedeckt: eine Team-Richtlinie mit gesperrten Datenschutz-Voreinstellungen, zentrale Steuerung und vertragliche Verschwiegenheitszusagen. Free und Pro sind für Einzelpersonen gedacht – ein AVV (Auftragsverarbeitungsvertrag, Art. 28 DSGVO) ist für Geschäftskunden in jedem Tarif enthalten, deckt aber § 203 StGB nicht ab. Free ist der dauerhaft kostenlose Tarif: 2.000 Wörter pro Woche mit allen Funktionen, Reset jeden Montag um 00:00 UTC, keine Kreditkarte. Jedes neue Konto startet mit einer 14-tägigen Pro-Testphase und läuft danach automatisch auf Free weiter. Zahlung für Pro über Mollie (europäischer Anbieter), Kreditkarte oder SEPA-Lastschrift. ### Häufige Fragen **Was passiert nach der 14-tägigen Testphase?** Nach Ablauf der Testphase läuft dein Konto automatisch im Free-Tarif weiter – 2.000 Wörter pro Woche mit allen Funktionen, ohne dass du etwas tun musst. Ein Upgrade auf Pro ist jederzeit möglich. Während der Testphase fallen keine Kosten an. **Kann ich jederzeit kündigen?** Ja, du kannst dein Abo jederzeit kündigen. Dein Zugang bleibt bis zum Ende des aktuellen Abrechnungszeitraums bestehen. **Welche Zahlungsmethoden akzeptiert ihr?** Wir akzeptieren alle gängigen Kreditkarten und SEPA-Lastschrift. Alle Zahlungen werden sicher über Mollie, einen europäischen Zahlungsanbieter, abgewickelt. **Gibt es einen kostenlosen Plan?** Ja. Jedes neue Konto startet mit einer 14-tägigen Pro-Testphase. Danach läuft dein Konto im Free-Tarif weiter: 2.000 Wörter pro Woche, alle Funktionen, keine Kreditkarte. Das Kontingent wird jeden Montag um 00:00 UTC zurückgesetzt, nicht genutzte Wörter verfallen. Ein Upgrade auf Pro ist jederzeit möglich. **Gibt es Studenten- oder Bildungsrabatte?** Ja! Kontaktiere uns mit einer gültigen Studenten-E-Mail und wir gewähren dir einen speziellen Rabatt auf dein Abo. **Wie funktioniert die Enterprise-Preisgestaltung?** Enterprise kostet 21,88 € netto pro Lizenz und Monat bei monatlicher Abrechnung oder 18,48 € netto pro Lizenz und Monat bei jährlicher Abrechnung (221,76 € pro Lizenz und Jahr), jeweils zzgl. MwSt. Du kannst Enterprise direkt online kaufen, per Rechnung bezahlen oder mit unserem Vertrieb sprechen, wenn du besondere Anforderungen hast. **Ich bin Arzt, Anwalt oder Steuerberater – welchen Tarif brauche ich?** Enterprise. Wenn du der Schweigepflicht unterliegst (Ärzte, Rechtsanwälte, Steuerberater, Notare, Psychotherapeuten und andere Berufsgeheimnisträger nach § 203 StGB), ist Enterprise der einzige Tarif, der das abdeckt: eine Team-Richtlinie mit gesperrten Datenschutz-Voreinstellungen, zentrale Steuerung und vertragliche Verschwiegenheitszusagen. Free und Pro sind für Einzelpersonen gedacht und decken § 203 StGB nicht ab. **Ist der AVV in Pro enthalten?** Ja. Ein AVV (Auftragsverarbeitungsvertrag nach Art. 28 DSGVO) ist für Geschäftskunden in jedem Tarif enthalten, auch in Free und Pro. Das Berufsgeheimnis nach § 203 StGB ist jedoch nur mit Enterprise abgedeckt. **Was ist Command und wann ist es verfügbar?** Mit Command steuerst du deinen Computer per Stimme, ohne das Diktieren zu verlassen: Ein kurzer Befehl wie „Drück Steuerung Z“ oder „Öffne Word“ wird ausgeführt statt geschrieben, ganze Sätze bleiben Text. Dazu kommen Menüpunkte jeder App per Name, eigene Befehle und mehrstufige Abläufe, der Import aus Dragon MyCommands und CommandMe, „Hey Sprecho“ für freihändiges Arbeiten, Text per Stimme markieren und korrigieren, Klicken und Maus per Stimme sowie Terminal Server Advanced für RDP- und Citrix-Sitzungen. Riskante Aktionen wie Schließen oder Löschen fragen vorher nach, und Admins verwalten Team-Befehlspakete und sperren Funktionen per Richtlinie. Command enthält alles aus Enterprise und kostet 44,99 € netto pro Lizenz und Monat oder 34,99 € netto pro Lizenz und Monat bei jährlicher Abrechnung, jeweils zzgl. MwSt. Command ist demnächst verfügbar – nimm Kontakt auf und wir melden uns vor dem Start bei dir. Aktuelle Preise: https://sprecho.ai/de/preise --- ## https://sprecho.ai/solutions/healthcare — Sprecho for Doctors — Dictate reports instead of typing > GDPR-compliant voice dictation for doctors and clinics. ICD codes, medication names, text snippets. 14-day free trial. ### Your last patient left at 5 PM. You're still at your desk at 7 PM. 23 reports, 8 referral letters, 4 prescriptions. With Sprecho, dictate everything directly into your practice software — during consultations, not after. #### Sound familiar? - **Documentation after hours** — Every evening at your desk typing reports, referral letters, and prescriptions. That's not medicine — that's typing. - **Patient data on US servers?** — Many dictation tools process audio data in the USA. At the next audit, you don't want to explain that. - **Medical terms aren't recognized** — ICD codes, medication names, anatomical terms — standard dictation tools fail on medical vocabulary. #### Features for healthcare - **Voice Dictation** — Dictate reports directly into your practice software, hands-free, during consultations. - **Medical Dictionary** — ICD codes, medication names, anatomical terms — Sprecho learns your vocabulary automatically. - **Text Snippets** — Standard findings with one shortcut. - **AI Formatting** — Automatic punctuation and grammar correction for your reports. - **Writing Styles** — Formal for referral letters, factual for findings — switch with one click. - **100+ Languages** — Patients with migration background? Sprecho understands any language. - **Quick Notes** — Capture thoughts between patients by voice. - **Practice Sharing** — Share medical dictionaries with your practice team. - **Patient Data Protected** — German servers, GDPR compliant, TÜV-certified DPO. #### Common questions from doctors **Does Sprecho work with my practice software?** Yes. Sprecho dictates anywhere your cursor is — EMR systems, Word, email, and any other application. **Is my patient data safe?** All data is stored in Germany. TÜV-certified DPO. Audio and text are deleted immediately after pasting. For medical professional secrecy (§ 203 StGB) there is the Enterprise plan. → Enterprise **Does Sprecho recognize medical terms?** Yes. Sprecho has a personal dictionary that automatically learns ICD codes, medication names, and anatomical terms. **What does Sprecho cost for my practice?** Single license from €10.92 net/month (€12.99 gross incl. VAT), practice licenses from €10.92 net/seat (€12.99 gross). 14-day free trial. --- ## https://sprecho.ai/solutions/legal — Sprecho for Lawyers — Dictate briefs instead of typing > GDPR-compliant voice dictation for law firms. Legal vocabulary, text snippets. Professional secrecy (§ 203 StGB) with the Enterprise plan. 14-day free trial. ### Your client pays €300/hour for your expertise. Not for you to type briefs. 495 hours per year at the keyboard instead of in the case file. With Sprecho, dictate directly into your legal software — with your legal vocabulary. #### Sound familiar? - **Hours typing briefs** — Each brief takes 45 minutes at the keyboard. 3 briefs per day × 220 work days = 495 hours per year. - **Attorney-client privilege at risk** — Sending client secrets to a US dictation tool can be a legal violation. You're personally liable. - **Legal terms get mangled** — Case numbers, statute references, legal terminology — standard tools turn them into unusable text. #### Features for legal - **Voice Dictation** — Dictate briefs directly into your legal software, with your legal vocabulary. - **Legal Dictionary** — Statute numbers, case references, legal terms — Sprecho learns your vocabulary automatically. - **Text Snippets** — Standard phrases and boilerplate with one shortcut. - **AI Formatting** — Automatic punctuation and grammar correction for professional briefs. - **Writing Styles** — Formal for courts, factual for clients — switch with one click. - **100+ Languages** — International clients? Sprecho understands any language. - **Quick Notes** — Capture notes during client meetings by voice. - **Firm Sharing** — Share legal dictionaries and text snippets with your team. - **Privilege Protected** — German servers, TÜV-certified DPO, professional secrecy (§ 203 StGB) with the Enterprise plan. #### Common questions from lawyers **Is Sprecho compatible with attorney-client privilege?** Yes, with the Enterprise plan: a team policy for professional secrecy with locked privacy defaults, central controls and contractual commitments. All data is stored in Germany; audio and text are deleted immediately after pasting. Free and Pro are for individuals (DPA for business customers on every plan). → Enterprise **Does Sprecho work with legal case management?** Yes. Sprecho dictates anywhere your cursor is — case management, Word, Outlook. **What does Sprecho cost for my firm?** Single license from €10.92 net/month (€12.99 gross incl. VAT), firm licenses from €10.92 net/seat (€12.99 gross). 14-day free trial. **Does Sprecho recognize legal terms?** Yes. Statute numbers, case references, and legal terminology are learned automatically. --- ## https://sprecho.ai/solutions/government — Sprecho for Government — Dictate protocols instead of typing > GDPR-compliant voice dictation for government offices. Text snippets for official documents. From €10.92 net/seat. ### 37 employees × 2 hours of typing per day. That's 74 lost work hours. Every day. Notices, memos, protocols, citizen inquiries — all typed. Sprecho lets your employees dictate instead of type. Setup in 5 minutes. #### Sound familiar? - **Typing protocols instead of governing** — After every council meeting, hours of typing protocols. Every employee types 2–3 hours per day. - **US tools in government?** — The data protection authority audits your IT. Citizen data on US servers — that's a compliance violation. - **Digitalization takes years** — The last IT project: 18 months, over budget, endless training. Nobody uses it. #### Features for government - **Voice Dictation** — Dictate notices and memos directly into your specialized application or DMS. - **Government Vocabulary** — File references, administrative terms, legal bases — Sprecho learns your vocabulary automatically. - **Text Snippets** — Standard notices and phrases with one shortcut. - **AI Formatting** — Automatic punctuation and grammar for professional notices. - **Writing Styles** — Formal for notices, neutral for memos — switch with one click. - **100+ Languages** — Citizens with migration background? Sprecho understands any language. - **Quick Notes** — Capture notes during citizen hours by voice. - **Office Sharing** — Share dictionaries and text snippets with the entire office. - **Citizen Data Protected** — German servers, GDPR compliant, TÜV-certified DPO. #### Common questions from government **Do we need a procurement process?** For individual licenses or small teams, usually not. Government licenses from €10.92 net/seat. **Does Sprecho work with our DMS?** Yes. Sprecho dictates anywhere the cursor is — in any specialized application, DMS, Word, or email. **How quickly can we start?** Setup takes 5 minutes per workstation. No IT project, no training. **Is citizen data safe with Sprecho?** All data is stored in Germany. TÜV-certified DPO. --- ## https://sprecho.ai/solutions/business — Sprecho for Business — 4x faster than typing > GDPR-compliant voice dictation for SMBs. Proposals, emails, reports. Team licenses from €10.92 net/seat. ### You're a business owner, not a typist. Yet you type 40 hours per month. Emails, proposals, reports — an entire work week per month goes to typing. Sprecho gives you that week back. #### Sound familiar? - **40 hours typing per month** — 40 words per minute typing vs. 150 speaking. You lose 2–3 hours every day. - **GDPR questionnaire from a client** — "Do you use cloud services with data processing outside the EU?" If your tool sends data to the US, the answer is: Yes. - **Every proposal from scratch** — Always the same phrases, always starting over. No system for recurring texts. #### Features for business - **Voice Dictation** — Dictate proposals, emails and reports directly into your software. - **Industry Vocabulary** — Company names, product names, jargon — Sprecho learns automatically. - **Text Snippets** — Standard proposals and greetings with one shortcut. - **AI Formatting** — Automatic punctuation and grammar for professional texts. - **Writing Styles** — Formal for clients, casual for the team — switch with one click. - **100+ Languages** — International clients? Sprecho understands any language. - **Quick Notes** — Capture ideas between meetings by voice. - **Team Sharing** — Share dictionaries and text snippets with your entire team. - **Customer Data Protected** — German servers, GDPR compliant, TÜV-certified DPO. #### Common questions from business owners **Does Sprecho work with my CRM?** Yes. Sprecho dictates anywhere your cursor is — CRM, email, Word, browser. **Are there team licenses?** Yes. From €10.92 net/seat for your team. Shared dictionaries included. **How safe is my customer data?** All data is stored in Germany. TÜV-certified DPO. **How quickly will I be productive?** Setup takes 2 minutes. Most users are 4x faster after the first day. --- ## https://sprecho.ai/solutions/tax-consultant — Sprecho for Tax Consultants — Dictate notices & appeals > GDPR-compliant voice dictation for tax firms. Direct in DATEV, Word, Outlook. Professional secrecy (§ 203 StGB) with the Enterprise plan. 14-day free trial. ### You're a tax consultant, not a typist. Yet you retype every single appeal. Reviewing notices, drafting appeals, client correspondence — hour after hour at the keyboard. With Sprecho, dictate everything directly into DATEV, Word and Outlook — 40% faster on tax notices and appeals. #### Sound familiar? - **Hours typing appeals** — Each appeal takes 30–45 minutes at the keyboard. During tax season that's hours of typing every day — instead of advising clients. - **Privilege and tax confidentiality** — Tax consultants are bound by professional secrecy too. Sending client data to a US dictation tool can be a legal violation. You're personally liable. - **Boilerplate phrases retyped every time** — Standard appeal reasoning, client letters, powers of attorney — same content, retyped from scratch each time. #### Features for tax-consultant - **Voice Dictation** — Dictate notices, appeals and client letters directly into DATEV, Word or Outlook. - **Tax Law Dictionary** — Statute numbers, tax IDs, court rulings, client names — Sprecho learns your vocabulary automatically. - **Text Snippets for Notices** — Standard appeal reasoning, powers of attorney, letters — one shortcut, done. - **AI Formatting** — Automatic punctuation and grammar for professional letters to tax authorities. - **Writing Styles** — Formal for tax authorities, factual for clients — switch with one click. - **100+ Languages** — International clients? Sprecho understands any language. - **Quick Notes** — Capture notes during client meetings by voice. - **Firm Sharing** — Share tax dictionaries and text snippets with your entire team. - **Client Data Protected** — German servers, TÜV-certified DPO, § 203 StGB with the Enterprise plan. #### Common questions from tax consultants **Does Sprecho work with DATEV?** Yes. Sprecho dictates anywhere your cursor is — DATEV (Eigenorganisation, Kanzlei-Rechnungswesen, ESt-Programm), Lexware, Addison, Word, Outlook, ELSTER. **Is Sprecho compatible with § 203 StGB and tax confidentiality?** Yes, with the Enterprise plan: a team policy for professional secrecy with locked privacy defaults, central controls and contractual commitments. All data is stored in Germany; audio and text are deleted immediately after pasting. TÜV-certified data protection officer. Free and Pro are for individuals (DPA for business customers on every plan). → Enterprise **What does Sprecho cost for my firm?** Single license from €10.92 net/month (€12.99 gross incl. VAT), firm licenses from €10.92 net/seat (€12.99 gross). 14-day free trial. **Does Sprecho recognize tax law terminology?** Yes. Statute numbers (§§ EStG, AO, UStG), tax IDs, BFH case numbers and client names are learned automatically. --- ## https://sprecho.ai/de/loesungen/aerzte — Sprecho für Ärzte — Befunde diktieren statt tippen > DSGVO-konformes Sprachdiktat für Ärzte und Praxen. ICD-Codes, Medikamentennamen, Textbausteine. 14 Tage kostenlos testen. ### Dein letzter Patient ging um 17 Uhr. Du sitzt um 19 Uhr noch am Rechner. 23 Befunde, 8 Arztbriefe, 4 Überweisungen. Mit Sprecho diktierst du alles direkt in deine Praxissoftware — während der Sprechstunde, nicht danach. #### Das kennst du als Arzt - **Dokumentation nach Feierabend** — Jeden Abend sitzt du am Rechner und tippst Befunde, Arztbriefe und Überweisungen. Das ist keine Medizin — das ist Schreibarbeit. - **Patientendaten auf US-Servern?** — Viele Diktiertools verarbeiten Audiodaten in den USA. Bei der nächsten KV-Prüfung willst du das nicht erklären müssen. - **Fachbegriffe werden nicht erkannt** — ICD-Codes, Medikamentennamen, anatomische Begriffe — Standard-Diktiertools scheitern an medizinischem Fachvokabular. #### Features for healthcare - **Sprachdiktat** — Diktiere Befunde direkt in deine Praxissoftware, freihändig, während der Sprechstunde. - **Medizinisches Wörterbuch** — ICD-Codes, Medikamentennamen, anatomische Begriffe — Sprecho lernt dein Fachvokabular automatisch. - **Textbausteine für Befunde** — "Befund unauffällig", "Medikation wie besprochen" — ein Kürzel genügt. - **KI-Formatierung** — Automatische Interpunktion und Grammatikkorrektur für deine Befunde. - **Schreibstile** — Formell für Arztbriefe, sachlich für Befunde — wechsle mit einem Klick. - **100+ Sprachen** — Patienten mit Migrationshintergrund? Sprecho versteht jede Sprache. - **Schnelle Notizen** — Erfasse Gedanken zwischen zwei Patienten per Stimme. - **Praxis-Sharing** — Teile medizinische Wörterbücher mit deinem Praxisteam. - **Patientendaten geschützt** — Deutsche Server, DSGVO-konform, TÜV-zertifizierter DSB. #### Häufige Fragen von Ärzten **Funktioniert Sprecho mit meiner Praxissoftware?** Ja. Sprecho diktiert überall, wo dein Cursor steht — CGM, mediatixx, Dampsoft, Word, E-Mail. **Sind meine Patientendaten sicher?** Alle Daten werden in Deutschland gespeichert. TÜV-zertifizierter DSB. Audio und Text werden sofort nach dem Einfügen gelöscht. Für die ärztliche Schweigepflicht (§ 203 StGB) gibt es den Enterprise-Plan. → Enterprise **Erkennt Sprecho medizinische Fachbegriffe?** Ja. Sprecho hat ein persönliches Wörterbuch, das ICD-Codes, Medikamentennamen und anatomische Begriffe automatisch lernt. **Was kostet Sprecho für meine Praxis?** Einzellizenz ab 10,92 € netto/Monat (12,99 € brutto inkl. MwSt.), Praxis-Lizenzen ab 10,92 € netto/Sitz (12,99 € brutto). 14 Tage kostenlos testen. --- ## https://sprecho.ai/de/loesungen/anwaelte — Sprecho für Anwälte — Schriftsätze diktieren statt tippen > DSGVO-konformes Sprachdiktat für Kanzleien. Schweigepflicht (§ 203 StGB) mit dem Enterprise-Plan. Juristisches Fachvokabular. 14 Tage kostenlos testen. ### Dein Mandant zahlt 300 € für deine Expertise. Nicht dafür, dass du Schriftsätze tippst. 495 Stunden pro Jahr am Keyboard statt in der Akte. Mit Sprecho diktierst du direkt in beA, Word und Outlook — mit deinem juristischen Fachvokabular. #### Das kennst du als Anwalt - **Stundenlang Schriftsätze tippen** — Jeder Schriftsatz dauert 45 Minuten am Keyboard. 3 Schriftsätze am Tag × 220 Arbeitstage = 495 Stunden pro Jahr. - **§ 203 StGB und Schweigepflicht** — Mandantengeheimnisse an ein US-Diktiertool zu schicken kann strafbar sein. Du haftest persönlich. - **Juristische Begriffe werden verstümmelt** — Aktenzeichen, Paragrafennummern, Fachbegriffe — Standard-Diktiertools machen daraus unbrauchbaren Text. #### Features for legal - **Sprachdiktat** — Diktiere Schriftsätze direkt in beA, Word oder Outlook, mit deinem juristischen Vokabular. - **Juristisches Wörterbuch** — Paragrafennummern, Aktenzeichen, Fachbegriffe — Sprecho lernt dein Vokabular automatisch. - **Textbausteine für die Kanzlei** — "Namens und in Vollmacht meiner Mandantschaft..." — ein Kürzel, fertig. - **KI-Formatierung** — Automatische Interpunktion und Grammatikkorrektur für professionelle Schriftsätze. - **Schreibstile** — Formell für Gerichte, sachlich für Mandanten — wechsle mit einem Klick. - **100+ Sprachen** — Internationale Mandanten? Sprecho versteht jede Sprache. - **Schnelle Notizen** — Erfasse Notizen während der Mandantenbesprechung per Stimme. - **Kanzlei-Sharing** — Teile juristische Wörterbücher und Textbausteine mit deinem Team. - **Mandantengeheimnis geschützt** — Deutsche Server, TÜV-zertifizierter DSB, § 203 StGB mit dem Enterprise-Plan. #### Häufige Fragen von Anwälten **Ist Sprecho mit § 203 StGB vereinbar?** Ja, mit dem Enterprise-Plan: Team-Richtlinie für Berufsgeheimnisse mit fest gesperrten Datenschutz-Standards, zentrale Kontrollen und vertragliche Zusagen. Alle Daten werden in Deutschland gespeichert, Audio und Text sofort nach dem Einfügen gelöscht. Free und Pro sind für Einzelpersonen (AVV für Geschäftskunden in jedem Plan). → Enterprise **Funktioniert Sprecho mit beA und RA-MICRO?** Ja. Sprecho diktiert überall, wo dein Cursor steht — beA, RA-MICRO, Word, Outlook. **Was kostet Sprecho für meine Kanzlei?** Einzellizenz ab 10,92 € netto/Monat (12,99 € brutto inkl. MwSt.), Kanzlei-Lizenzen ab 10,92 € netto/Sitz (12,99 € brutto). 14 Tage kostenlos testen. **Erkennt Sprecho juristische Fachbegriffe?** Ja. Paragrafennummern, Aktenzeichen und juristische Fachbegriffe werden automatisch gelernt. --- ## https://sprecho.ai/de/loesungen/verwaltung — Sprecho für Verwaltung — Protokolle diktieren statt tippen > DSGVO-konformes Sprachdiktat für Behörden und Verwaltung. Textbausteine für Bescheide. Ab 10,92 € netto/Sitz. ### 37 Sachbearbeiter × 2 Stunden Tippen pro Tag. Das sind 74 verlorene Arbeitsstunden. Jeden Tag. Bescheide, Vermerke, Protokolle, Bürgeranfragen — alles wird getippt. Sprecho lässt deine Mitarbeiter diktieren statt tippen. Installation in 5 Minuten. #### Das kennst du in der Verwaltung - **Protokolle tippen statt regieren** — Nach jeder Ratssitzung stundenlang Protokoll tippen. Jeder Sachbearbeiter tippt 2–3 Stunden am Tag. - **US-Tools in der Behörde?** — Der Landesbeauftragte prüft eure IT. Bürgerdaten auf US-Servern — das ist ein Compliance-Verstoß. - **Digitalisierung dauert Jahre** — Das letzte IT-Projekt: 18 Monate, Budget überzogen, Schulungen ohne Ende. Keiner nutzt es. #### Features for government - **Sprachdiktat** — Diktiere Bescheide und Vermerke direkt in euer Fachverfahren oder DMS. - **Verwaltungsvokabular** — Aktenzeichen, Behördenbegriffe, Rechtsgrundlagen — Sprecho lernt euer Vokabular automatisch. - **Textbausteine für Bescheide** — Standardbescheide und Formulierungen mit einem Kürzel einfügen. - **KI-Formatierung** — Automatische Interpunktion und Grammatik für professionelle Bescheide. - **Schreibstile** — Formell für Bescheide, neutral für Vermerke — wechsle mit einem Klick. - **100+ Sprachen** — Bürger mit Migrationshintergrund? Sprecho versteht jede Sprache. - **Schnelle Notizen** — Erfasse Notizen während der Sprechstunde per Stimme. - **Amt-Sharing** — Teile Wörterbücher und Textbausteine mit dem gesamten Amt. - **Bürgerdaten geschützt** — Deutsche Server, DSGVO-konform, TÜV-zertifizierter DSB. #### Häufige Fragen aus der Verwaltung **Brauchen wir eine Ausschreibung?** Für Einzellizenzen oder kleine Teams in der Regel nicht. Ab 10,92 € netto/Sitz sind Verwaltungslizenzen verfügbar. **Funktioniert Sprecho mit unserem DMS?** Ja. Sprecho diktiert überall, wo der Cursor steht — in jedem Fachverfahren, DMS, Word oder E-Mail. **Wie schnell können wir starten?** Installation dauert 5 Minuten pro Arbeitsplatz. Kein IT-Projekt, keine Schulung. **Sind Bürgerdaten bei Sprecho sicher?** Alle Daten werden in Deutschland gespeichert. TÜV-zertifizierter DSB. --- ## https://sprecho.ai/de/loesungen/unternehmen — Sprecho für Unternehmen — 4x schneller als Tippen > DSGVO-konformes Sprachdiktat für KMU. Angebote, E-Mails, Berichte diktieren. Ab 10,92 € netto/Sitz. ### Du bist Geschäftsführer, nicht Sekretär. Trotzdem tippst du 40 Stunden im Monat. E-Mails, Angebote, Berichte — eine komplette Arbeitswoche pro Monat geht fürs Tippen drauf. Sprecho gibt dir diese Woche zurück. #### Das kennst du als Unternehmer - **40 Stunden Tippen pro Monat** — 40 Wörter pro Minute tippen vs. 150 Wörter pro Minute sprechen. Du verlierst 2–3 Stunden pro Tag. - **DSGVO-Fragebogen vom Kunden** — "Nutzen Sie Cloud-Dienste mit Datenverarbeitung außerhalb der EU?" Wenn dein Tool Daten in die USA schickt, ist die Antwort: Ja. - **Jedes Angebot von Null anfangen** — Immer die gleichen Formulierungen, immer von vorne. Kein System für wiederkehrende Texte. #### Features for business - **Sprachdiktat** — Diktiere Angebote, E-Mails und Berichte direkt in deine Software. - **Branchenvokabular** — Firmennamen, Produktnamen, Fachjargon — Sprecho lernt automatisch. - **Textbausteine für Angebote** — Standardangebote und Grußformeln mit einem Kürzel einfügen. - **KI-Formatierung** — Automatische Interpunktion und Grammatik für professionelle Texte. - **Schreibstile** — Formell für Kunden, locker für das Team — wechsle mit einem Klick. - **100+ Sprachen** — Internationale Kunden? Sprecho versteht jede Sprache. - **Schnelle Notizen** — Erfasse Ideen zwischen Meetings per Stimme. - **Team-Sharing** — Teile Wörterbücher und Textbausteine mit deinem ganzen Team. - **Kundendaten geschützt** — Deutsche Server, DSGVO-konform, TÜV-zertifizierter DSB. #### Häufige Fragen von Unternehmern **Funktioniert Sprecho mit meinem CRM?** Ja. Sprecho diktiert überall, wo dein Cursor steht — CRM, E-Mail, Word, Browser. **Gibt es Team-Lizenzen?** Ja. Ab 10,92 € netto/Sitz für dein Team. Gemeinsame Wörterbücher inklusive. **Wie sicher sind meine Kundendaten?** Alle Daten werden in Deutschland gespeichert. TÜV-zertifizierter DSB. **Wie schnell bin ich produktiv?** Installation dauert 2 Minuten. Die meisten Nutzer sind nach dem ersten Tag 4x schneller. --- ## https://sprecho.ai/de/loesungen/steuerberater — Sprecho für Steuerberater — Bescheide & Einsprüche diktieren > DSGVO-konformes Sprachdiktat für Steuerkanzleien. Direkt in DATEV, Word und Outlook. Schweigepflicht (§ 203 StGB) mit dem Enterprise-Plan. 14 Tage kostenlos. ### Du bist Steuerberater, keine Schreibkraft. Trotzdem tippst du jeden Bescheid neu. Bescheide prüfen, Einsprüche begründen, Mandantenkorrespondenz tippen — Stunde um Stunde. Mit Sprecho diktierst du alles direkt in DATEV, Word und Outlook — 40 % schneller bei Bescheiden und Einsprüchen. #### Das kennst du als Steuerberater - **Stundenlang Einsprüche tippen** — Jeder Einspruch dauert 30–45 Minuten am Keyboard. In der Saison sind das mehrere Stunden Tipparbeit pro Tag — statt Mandanten zu beraten. - **§ 203 StGB und Steuergeheimnis** — Auch Steuerberater unterliegen der Schweigepflicht. Mandantendaten an ein US-Diktiertool zu schicken kann strafbar sein. Du haftest persönlich. - **Wiederkehrende Formulierungen jedes Mal von Null** — Standardbegründungen für Einsprüche, Mandantenanschreiben, Vollmachten — immer das gleiche, immer von vorne tippen. #### Features for tax-consultant - **Sprachdiktat** — Diktiere Bescheide, Einsprüche und Mandantenanschreiben direkt in DATEV, Word oder Outlook. - **Steuerrechtliches Wörterbuch** — Paragrafennummern, Steuernummern, BFH-Urteile, Mandantennamen — Sprecho lernt dein Vokabular automatisch. - **Textbausteine für Bescheide** — Standardbegründungen für Einsprüche, Vollmachten, Anschreiben — ein Kürzel, fertig. - **KI-Formatierung** — Automatische Interpunktion und Grammatik für professionelle Schreiben an Finanzbehörden. - **Schreibstile** — Formell für Finanzämter, sachlich für Mandanten — wechsle mit einem Klick. - **100+ Sprachen** — Internationale Mandanten? Sprecho versteht jede Sprache. - **Schnelle Notizen** — Erfasse Notizen während der Mandantengespräche per Stimme. - **Kanzlei-Sharing** — Teile steuerrechtliche Wörterbücher und Textbausteine mit deinem ganzen Team. - **Mandantendaten geschützt** — Deutsche Server, TÜV-zertifizierter DSB, § 203 StGB mit dem Enterprise-Plan. #### Häufige Fragen von Steuerberatern **Funktioniert Sprecho mit DATEV?** Ja. Sprecho diktiert überall, wo dein Cursor steht — DATEV (Eigenorganisation, Kanzlei-Rechnungswesen, ESt-Programm), Lexware, Addison, Word, Outlook, ELSTER. **Ist Sprecho mit § 203 StGB und dem Steuergeheimnis vereinbar?** Ja, mit dem Enterprise-Plan: Team-Richtlinie für Berufsgeheimnisse mit fest gesperrten Datenschutz-Standards, zentrale Kontrollen und vertragliche Zusagen. Alle Daten werden in Deutschland gespeichert, Audio und Text sofort nach dem Einfügen gelöscht. TÜV-zertifizierter Datenschutzbeauftragter. Free und Pro sind für Einzelpersonen (AVV für Geschäftskunden in jedem Plan). → Enterprise **Was kostet Sprecho für meine Kanzlei?** Einzellizenz ab 10,92 € netto/Monat (12,99 € brutto inkl. MwSt.), Kanzlei-Lizenzen ab 10,92 € netto/Sitz (12,99 € brutto). 14 Tage kostenlos testen. **Erkennt Sprecho steuerrechtliche Fachbegriffe?** Ja. Paragrafennummern (§§ EStG, AO, UStG), Steuernummern, BFH-Aktenzeichen und Mandantennamen werden automatisch gelernt. --- ## https://sprecho.ai/compare — Compare Sprecho to every voice-dictation tool — Wispr Flow, Superwhisper, Voicely & more > Side-by-side comparisons of Sprecho against Wispr Flow, Superwhisper, Voicely, MacWhisper, Dragon, Otter.ai and Apple Dictation. Free forever with 2,000 words/week. DSGVO-compliant. Built in Germany. ### Voice dictation compared — Sprecho vs Wispr Flow, Superwhisper & more Sprecho is voice dictation software for Windows, macOS and Linux — the DSGVO-compliant alternative to Wispr Flow, Superwhisper and Voicely, built in Germany. Free forever — 2,000 words per week, no credit card. --- ## https://sprecho.ai/compare/sprecho-vs-wispr-flow — Sprecho vs Wispr Flow > Honest side-by-side comparison of Sprecho (GDPR-compliant, EU-hosted, German Fachsprache) against Wispr Flow. **Platform support** - Windows: Yes - macOS: Yes - Linux: No - iOS: Yes - Android: Yes **Privacy & DSGVO** - Servers in the EU: No - Not subject to US CLOUD Act: No - Does not take screenshots: Partial - DPA / AVV available: Partial - Public subprocessor list: Partial **German language quality** - High German accuracy: Partial - Professional vocabulary (DATEV, RA-MICRO, Medistar): No - German compound words: Partial - Regional dialects: Partial **Pricing** - Free tier — no credit card: Yes - Yearly discount: Yes - Enterprise scaling (volume discounts): No - VAT-inclusive pricing (EUR): No **Dictation features** - Personal dictionary (learns your words): Yes - Shared team dictionary: Yes - Voice snippets / shortcuts: Yes - Voice commands (new paragraph, comma, …): Yes - Standalone voice notes: Yes - Multiple output styles per app (formal / casual / …): Yes - Caps Lock → UPPERCASE dictation: No --- ## https://sprecho.ai/compare/sprecho-vs-superwhisper — Sprecho vs Superwhisper > Honest side-by-side comparison of Sprecho (GDPR-compliant, EU-hosted, German Fachsprache) against Superwhisper. **Platform support** - Windows: Yes - macOS: Yes - Linux: No - iOS: Yes - Android: No **Privacy & DSGVO** - Servers in the EU: Partial - Not subject to US CLOUD Act: Partial - Does not take screenshots: Yes - DPA / AVV available: No - Public subprocessor list: No **German language quality** - High German accuracy: Partial - Professional vocabulary (DATEV, RA-MICRO, Medistar): No - German compound words: Partial - Regional dialects: No **Pricing** - Free tier — no credit card: Yes - Yearly discount: Yes - Enterprise scaling (volume discounts): Yes - VAT-inclusive pricing (EUR): No **Dictation features** - Personal dictionary (learns your words): Yes - Shared team dictionary: No - Voice snippets / shortcuts: Partial - Voice commands (new paragraph, comma, …): No - Standalone voice notes: No - Multiple output styles per app (formal / casual / …): Yes - Caps Lock → UPPERCASE dictation: No --- ## https://sprecho.ai/compare/sprecho-vs-voicely — Sprecho vs Voicely > Honest side-by-side comparison of Sprecho (GDPR-compliant, EU-hosted, German Fachsprache) against Voicely. **Platform support** - Windows: Yes - macOS: Yes - Linux: No - iOS: No - Android: No **Privacy & DSGVO** - Servers in the EU: Partial - Not subject to US CLOUD Act: Partial - Does not take screenshots: Yes - DPA / AVV available: Partial - Public subprocessor list: Partial **German language quality** - High German accuracy: Yes - Professional vocabulary (DATEV, RA-MICRO, Medistar): No - German compound words: Yes - Regional dialects: No **Pricing** - Free tier — no credit card: Yes - Yearly discount: Yes - Enterprise scaling (volume discounts): Partial - VAT-inclusive pricing (EUR): No **Dictation features** - Personal dictionary (learns your words): Yes - Shared team dictionary: Yes - Voice snippets / shortcuts: Partial - Voice commands (new paragraph, comma, …): No - Standalone voice notes: No - Multiple output styles per app (formal / casual / …): Partial - Caps Lock → UPPERCASE dictation: No --- ## https://sprecho.ai/compare/sprecho-vs-macwhisper — Sprecho vs MacWhisper > Honest side-by-side comparison of Sprecho (GDPR-compliant, EU-hosted, German Fachsprache) against MacWhisper. **Platform support** - Windows: No - macOS: Yes - Linux: No - iOS: No - Android: No **Privacy & DSGVO** - Servers in the EU: Yes - Not subject to US CLOUD Act: Yes - Does not take screenshots: Yes - DPA / AVV available: No - Public subprocessor list: No **German language quality** - High German accuracy: Partial - Professional vocabulary (DATEV, RA-MICRO, Medistar): No - German compound words: Partial - Regional dialects: No **Pricing** - Free tier — no credit card: Yes - Yearly discount: No - Enterprise scaling (volume discounts): No - VAT-inclusive pricing (EUR): No **Dictation features** - Personal dictionary (learns your words): Yes - Shared team dictionary: No - Voice snippets / shortcuts: No - Voice commands (new paragraph, comma, …): No - Standalone voice notes: Yes - Multiple output styles per app (formal / casual / …): Partial - Caps Lock → UPPERCASE dictation: No --- ## https://sprecho.ai/compare/sprecho-vs-dragon — Sprecho vs Dragon Professional > Honest side-by-side comparison of Sprecho (GDPR-compliant, EU-hosted, German Fachsprache) against Dragon Professional. **Platform support** - Windows: Yes - macOS: No - Linux: No - iOS: No - Android: No **Privacy & DSGVO** - Servers in the EU: Partial - Not subject to US CLOUD Act: Partial - Does not take screenshots: Yes - DPA / AVV available: Partial - Public subprocessor list: Partial **German language quality** - High German accuracy: Yes - Professional vocabulary (DATEV, RA-MICRO, Medistar): Yes - German compound words: Yes - Regional dialects: Partial **Pricing** - Free tier — no credit card: No - Yearly discount: Partial - Enterprise scaling (volume discounts): Yes - VAT-inclusive pricing (EUR): No **Dictation features** - Personal dictionary (learns your words): Yes - Shared team dictionary: Yes - Voice snippets / shortcuts: Yes - Voice commands (new paragraph, comma, …): Yes - Standalone voice notes: Partial - Multiple output styles per app (formal / casual / …): No - Caps Lock → UPPERCASE dictation: No --- ## https://sprecho.ai/compare/sprecho-vs-otter-ai — Sprecho vs Otter.ai > Honest side-by-side comparison of Sprecho (GDPR-compliant, EU-hosted, German Fachsprache) against Otter.ai. **Platform support** - Windows: Yes - macOS: Yes - Linux: No - iOS: Yes - Android: Yes **Privacy & DSGVO** - Servers in the EU: No - Not subject to US CLOUD Act: No - Does not take screenshots: Yes - DPA / AVV available: No - Public subprocessor list: Yes **German language quality** - High German accuracy: Partial - Professional vocabulary (DATEV, RA-MICRO, Medistar): No - German compound words: Partial - Regional dialects: No **Pricing** - Free tier — no credit card: Yes - Yearly discount: Yes - Enterprise scaling (volume discounts): Partial - VAT-inclusive pricing (EUR): No **Dictation features** - Personal dictionary (learns your words): Yes - Shared team dictionary: Yes - Voice snippets / shortcuts: No - Voice commands (new paragraph, comma, …): Partial - Standalone voice notes: Yes - Multiple output styles per app (formal / casual / …): No - Caps Lock → UPPERCASE dictation: No --- ## https://sprecho.ai/compare/sprecho-vs-apple-dictation — Sprecho vs Apple Dictation > Honest side-by-side comparison of Sprecho (GDPR-compliant, EU-hosted, German Fachsprache) against Apple Dictation. **Platform support** - Windows: No - macOS: Yes - Linux: No - iOS: Yes - Android: No **Privacy & DSGVO** - Servers in the EU: Partial - Not subject to US CLOUD Act: Partial - Does not take screenshots: Yes - DPA / AVV available: No - Public subprocessor list: No **German language quality** - High German accuracy: Partial - Professional vocabulary (DATEV, RA-MICRO, Medistar): No - German compound words: Partial - Regional dialects: Partial **Pricing** - Free tier — no credit card: Yes - Yearly discount: Yes - Enterprise scaling (volume discounts): No - VAT-inclusive pricing (EUR): Yes **Dictation features** - Personal dictionary (learns your words): No - Shared team dictionary: No - Voice snippets / shortcuts: No - Voice commands (new paragraph, comma, …): Partial - Standalone voice notes: No - Multiple output styles per app (formal / casual / …): No - Caps Lock → UPPERCASE dictation: No --- ## https://sprecho.ai/de/vergleich — Sprecho im Vergleich — Wispr Flow, Superwhisper, Voicely & mehr > Direkte Vergleiche von Sprecho mit Wispr Flow, Superwhisper, Voicely, MacWhisper, Dragon, Otter.ai und Apple Diktat. Kostenlos für immer — 2.000 Wörter/Woche, ohne Kreditkarte. DSGVO-konform. Made in Germany. ### Diktiersoftware im Vergleich — Sprecho vs. Wispr Flow, Superwhisper & mehr Sprecho ist die Diktiersoftware für Windows, macOS und Linux — die DSGVO-konforme Alternative zu Wispr Flow, Superwhisper und Voicely, made in Germany. Kostenlos für immer — 2.000 Wörter pro Woche, ohne Kreditkarte. --- ## https://sprecho.ai/de/vergleich/sprecho-vs-wispr-flow — Sprecho vs Wispr Flow > Honest side-by-side comparison of Sprecho (GDPR-compliant, EU-hosted, German Fachsprache) against Wispr Flow. **Plattform-Unterstützung** - Windows: Ja - macOS: Ja - Linux: Nein - iOS: Ja - Android: Ja **Datenschutz & DSGVO** - Server in der EU: Nein - Nicht dem US CLOUD Act unterworfen: Nein - Erstellt keine Screenshots: Teilweise - AVV / Auftragsverarbeitungsvertrag verfügbar: Teilweise - Öffentliche Liste der Unter­auftragsverarbeiter: Teilweise **Qualität deutscher Sprache** - Hohe Erkennungsgenauigkeit auf Deutsch: Teilweise - Fachsprache (DATEV, RA-MICRO, Medistar): Nein - Deutsche Komposita: Teilweise - Regionale Dialekte: Teilweise **Preise** - Free-Plan — ohne Kreditkarte: Ja - Jahresrabatt: Ja - Enterprise-Skalierung (Mengenrabatte): Nein - Brutto-Preise inkl. MwSt. (EUR): Nein **Diktier-Funktionen** - Persönliches Wörterbuch (lernt deine Begriffe): Ja - Geteiltes Team-Wörterbuch: Ja - Sprach-Snippets / Kürzel: Ja - Sprachbefehle (neuer Absatz, Komma, …): Ja - Eigenständige Sprachnotizen: Ja - Mehrere Ausgabestile pro App (formal / locker / …): Ja - Caps Lock → GROSSBUCHSTABEN-Diktat: Nein --- ## https://sprecho.ai/de/vergleich/sprecho-vs-superwhisper — Sprecho vs Superwhisper > Honest side-by-side comparison of Sprecho (GDPR-compliant, EU-hosted, German Fachsprache) against Superwhisper. **Plattform-Unterstützung** - Windows: Ja - macOS: Ja - Linux: Nein - iOS: Ja - Android: Nein **Datenschutz & DSGVO** - Server in der EU: Teilweise - Nicht dem US CLOUD Act unterworfen: Teilweise - Erstellt keine Screenshots: Ja - AVV / Auftragsverarbeitungsvertrag verfügbar: Nein - Öffentliche Liste der Unter­auftragsverarbeiter: Nein **Qualität deutscher Sprache** - Hohe Erkennungsgenauigkeit auf Deutsch: Teilweise - Fachsprache (DATEV, RA-MICRO, Medistar): Nein - Deutsche Komposita: Teilweise - Regionale Dialekte: Nein **Preise** - Free-Plan — ohne Kreditkarte: Ja - Jahresrabatt: Ja - Enterprise-Skalierung (Mengenrabatte): Ja - Brutto-Preise inkl. MwSt. (EUR): Nein **Diktier-Funktionen** - Persönliches Wörterbuch (lernt deine Begriffe): Ja - Geteiltes Team-Wörterbuch: Nein - Sprach-Snippets / Kürzel: Teilweise - Sprachbefehle (neuer Absatz, Komma, …): Nein - Eigenständige Sprachnotizen: Nein - Mehrere Ausgabestile pro App (formal / locker / …): Ja - Caps Lock → GROSSBUCHSTABEN-Diktat: Nein --- ## https://sprecho.ai/de/vergleich/sprecho-vs-voicely — Sprecho vs Voicely > Honest side-by-side comparison of Sprecho (GDPR-compliant, EU-hosted, German Fachsprache) against Voicely. **Plattform-Unterstützung** - Windows: Ja - macOS: Ja - Linux: Nein - iOS: Nein - Android: Nein **Datenschutz & DSGVO** - Server in der EU: Teilweise - Nicht dem US CLOUD Act unterworfen: Teilweise - Erstellt keine Screenshots: Ja - AVV / Auftragsverarbeitungsvertrag verfügbar: Teilweise - Öffentliche Liste der Unter­auftragsverarbeiter: Teilweise **Qualität deutscher Sprache** - Hohe Erkennungsgenauigkeit auf Deutsch: Ja - Fachsprache (DATEV, RA-MICRO, Medistar): Nein - Deutsche Komposita: Ja - Regionale Dialekte: Nein **Preise** - Free-Plan — ohne Kreditkarte: Ja - Jahresrabatt: Ja - Enterprise-Skalierung (Mengenrabatte): Teilweise - Brutto-Preise inkl. MwSt. (EUR): Nein **Diktier-Funktionen** - Persönliches Wörterbuch (lernt deine Begriffe): Ja - Geteiltes Team-Wörterbuch: Ja - Sprach-Snippets / Kürzel: Teilweise - Sprachbefehle (neuer Absatz, Komma, …): Nein - Eigenständige Sprachnotizen: Nein - Mehrere Ausgabestile pro App (formal / locker / …): Teilweise - Caps Lock → GROSSBUCHSTABEN-Diktat: Nein --- ## https://sprecho.ai/de/vergleich/sprecho-vs-macwhisper — Sprecho vs MacWhisper > Honest side-by-side comparison of Sprecho (GDPR-compliant, EU-hosted, German Fachsprache) against MacWhisper. **Plattform-Unterstützung** - Windows: Nein - macOS: Ja - Linux: Nein - iOS: Nein - Android: Nein **Datenschutz & DSGVO** - Server in der EU: Ja - Nicht dem US CLOUD Act unterworfen: Ja - Erstellt keine Screenshots: Ja - AVV / Auftragsverarbeitungsvertrag verfügbar: Nein - Öffentliche Liste der Unter­auftragsverarbeiter: Nein **Qualität deutscher Sprache** - Hohe Erkennungsgenauigkeit auf Deutsch: Teilweise - Fachsprache (DATEV, RA-MICRO, Medistar): Nein - Deutsche Komposita: Teilweise - Regionale Dialekte: Nein **Preise** - Free-Plan — ohne Kreditkarte: Ja - Jahresrabatt: Nein - Enterprise-Skalierung (Mengenrabatte): Nein - Brutto-Preise inkl. MwSt. (EUR): Nein **Diktier-Funktionen** - Persönliches Wörterbuch (lernt deine Begriffe): Ja - Geteiltes Team-Wörterbuch: Nein - Sprach-Snippets / Kürzel: Nein - Sprachbefehle (neuer Absatz, Komma, …): Nein - Eigenständige Sprachnotizen: Ja - Mehrere Ausgabestile pro App (formal / locker / …): Teilweise - Caps Lock → GROSSBUCHSTABEN-Diktat: Nein --- ## https://sprecho.ai/de/vergleich/sprecho-vs-dragon — Sprecho vs Dragon Professional > Honest side-by-side comparison of Sprecho (GDPR-compliant, EU-hosted, German Fachsprache) against Dragon Professional. **Plattform-Unterstützung** - Windows: Ja - macOS: Nein - Linux: Nein - iOS: Nein - Android: Nein **Datenschutz & DSGVO** - Server in der EU: Teilweise - Nicht dem US CLOUD Act unterworfen: Teilweise - Erstellt keine Screenshots: Ja - AVV / Auftragsverarbeitungsvertrag verfügbar: Teilweise - Öffentliche Liste der Unter­auftragsverarbeiter: Teilweise **Qualität deutscher Sprache** - Hohe Erkennungsgenauigkeit auf Deutsch: Ja - Fachsprache (DATEV, RA-MICRO, Medistar): Ja - Deutsche Komposita: Ja - Regionale Dialekte: Teilweise **Preise** - Free-Plan — ohne Kreditkarte: Nein - Jahresrabatt: Teilweise - Enterprise-Skalierung (Mengenrabatte): Ja - Brutto-Preise inkl. MwSt. (EUR): Nein **Diktier-Funktionen** - Persönliches Wörterbuch (lernt deine Begriffe): Ja - Geteiltes Team-Wörterbuch: Ja - Sprach-Snippets / Kürzel: Ja - Sprachbefehle (neuer Absatz, Komma, …): Ja - Eigenständige Sprachnotizen: Teilweise - Mehrere Ausgabestile pro App (formal / locker / …): Nein - Caps Lock → GROSSBUCHSTABEN-Diktat: Nein --- ## https://sprecho.ai/de/vergleich/sprecho-vs-otter-ai — Sprecho vs Otter.ai > Honest side-by-side comparison of Sprecho (GDPR-compliant, EU-hosted, German Fachsprache) against Otter.ai. **Plattform-Unterstützung** - Windows: Ja - macOS: Ja - Linux: Nein - iOS: Ja - Android: Ja **Datenschutz & DSGVO** - Server in der EU: Nein - Nicht dem US CLOUD Act unterworfen: Nein - Erstellt keine Screenshots: Ja - AVV / Auftragsverarbeitungsvertrag verfügbar: Nein - Öffentliche Liste der Unter­auftragsverarbeiter: Ja **Qualität deutscher Sprache** - Hohe Erkennungsgenauigkeit auf Deutsch: Teilweise - Fachsprache (DATEV, RA-MICRO, Medistar): Nein - Deutsche Komposita: Teilweise - Regionale Dialekte: Nein **Preise** - Free-Plan — ohne Kreditkarte: Ja - Jahresrabatt: Ja - Enterprise-Skalierung (Mengenrabatte): Teilweise - Brutto-Preise inkl. MwSt. (EUR): Nein **Diktier-Funktionen** - Persönliches Wörterbuch (lernt deine Begriffe): Ja - Geteiltes Team-Wörterbuch: Ja - Sprach-Snippets / Kürzel: Nein - Sprachbefehle (neuer Absatz, Komma, …): Teilweise - Eigenständige Sprachnotizen: Ja - Mehrere Ausgabestile pro App (formal / locker / …): Nein - Caps Lock → GROSSBUCHSTABEN-Diktat: Nein --- ## https://sprecho.ai/de/vergleich/sprecho-vs-apple-diktat — Sprecho vs Apple Dictation > Honest side-by-side comparison of Sprecho (GDPR-compliant, EU-hosted, German Fachsprache) against Apple Dictation. **Plattform-Unterstützung** - Windows: Nein - macOS: Ja - Linux: Nein - iOS: Ja - Android: Nein **Datenschutz & DSGVO** - Server in der EU: Teilweise - Nicht dem US CLOUD Act unterworfen: Teilweise - Erstellt keine Screenshots: Ja - AVV / Auftragsverarbeitungsvertrag verfügbar: Nein - Öffentliche Liste der Unter­auftragsverarbeiter: Nein **Qualität deutscher Sprache** - Hohe Erkennungsgenauigkeit auf Deutsch: Teilweise - Fachsprache (DATEV, RA-MICRO, Medistar): Nein - Deutsche Komposita: Teilweise - Regionale Dialekte: Teilweise **Preise** - Free-Plan — ohne Kreditkarte: Ja - Jahresrabatt: Ja - Enterprise-Skalierung (Mengenrabatte): Nein - Brutto-Preise inkl. MwSt. (EUR): Ja **Diktier-Funktionen** - Persönliches Wörterbuch (lernt deine Begriffe): Nein - Geteiltes Team-Wörterbuch: Nein - Sprach-Snippets / Kürzel: Nein - Sprachbefehle (neuer Absatz, Komma, …): Teilweise - Eigenständige Sprachnotizen: Nein - Mehrere Ausgabestile pro App (formal / locker / …): Nein - Caps Lock → GROSSBUCHSTABEN-Diktat: Nein --- ## https://sprecho.ai/blog/wispr-flow-lawyers-doctors-tax-advisors-germany-hipaa-is-not-203-stgb — Wispr Flow for Lawyers, Doctors and Tax Advisors in Germany: Why HIPAA Is Not § 203 StGB (2026) > Is Wispr Flow safe for German lawyers, doctors and tax advisors? What HIPAA, SOC 2 and ISO 27001 prove, what § 203 StGB demands, plus a DPO checklist. *Published: 2026-11-20T18:03:47.219000Z* *Tags: wispr-flow, legal, healthcare, gdpr, germany, privacy, comparison* **Wispr Flow for lawyers, doctors and tax advisors** looks like an easy decision on paper: the US dictation app advertises a signed HIPAA Business Associate Agreement, a SOC 2 Type II report and an ISO/IEC 27001:2022 certificate. For a German Kanzlei, Arztpraxis or Steuerkanzlei, though, none of those three documents answers the question that actually carries criminal liability: does using the tool count as an unauthorised disclosure of a client's or patient's secret under **§ 203 StGB**? *Last verified: October 2026.* This guide explains what each certificate really attests, what German professional secrecy law requires word for word, and how Wispr Flow's own published facts line up against those requirements. We stay fair: Wispr Flow is fast, polished and well funded, and its US compliance work is real. It simply answers a different legal system. - **Short answer:** HIPAA, SOC 2 and ISO 27001 describe US health-privacy duties and security controls. § 203 StGB, § 43e BRAO and § 62a StBerG ask who may learn a secret, under which written undertaking, and whether protection abroad is comparable to Germany. - **Wispr Flow's position:** by its own words a cloud service that "processes and stores customer data in the United States", operated by Wispr AI, Inc., a Delaware corporation. - **What that means:** not an automatic ban, but the full burden of proof sits with you as the professional, and most data protection officers will not sign it off for mandate or patient content. - **The alternative:** an EU provider with a public AVV, a public sub-processor list and a separate confidentiality undertaking, which is how [Sprecho for law firms](https://sprecho.ai/solutions/legal) is set up. ## The short answer: HIPAA compliance does not make Wispr Flow § 203-ready **No certificate Wispr Flow holds is designed to prove compliance with German professional secrecy law**, because none of them was written for it. They are useful evidence of good security practice, and a data protection officer will still ask for them, but they are inputs to your assessment, not the result. - **HIPAA** is a US federal law from 1996 that governs "covered entities" such as US health plans and providers and their "business associates". A German practice is not a covered entity, so a BAA gives it no rights under German law. - **SOC 2 Type II** is an auditor's report on whether a company's controls worked over a period. It does not say where data may lawfully go or who may read it under foreign law. - **ISO/IEC 27001** certifies that an information security management system exists and is maintained. It says nothing about the CLOUD Act or about § 203 StGB. - **§ 203 StGB** punishes the unauthorised disclosure of a secret with up to one year in prison or a fine, and since 2017 it also makes you liable if you fail to bind a service provider to secrecy. ## What Wispr Flow is and what it publishes about compliance **Wispr Flow is an AI voice dictation app by Wispr AI, Inc. (San Francisco) that turns speech into formatted text in any app on Mac, Windows, iPhone and Android.** Its compliance pages are unusually detailed for a consumer-grade tool, which is why it keeps appearing on shortlists of German professionals. Here is what Wispr publishes, quoted where the wording matters (all sources accessed on 5 October 2026). - **Processing location:** "Wispr Flow is a cloud service, not an on-premise deployment, and it processes and stores customer data in the United States." ([Wispr Flow security and compliance FAQ](https://docs.wisprflow.ai/articles/3467817258-security-and-compliance-faq)) - **Certifications:** the Trust Center "provides Wispr's SOC 2 Type II report, the ISO/IEC 27001:2022 certificate, the earlier SOC 2 Type I report". The reports sit behind an access request on the [Wispr Trust Center](https://trust.wispr.ai/). - **Audit scope:** "Wispr Flow dictation is the in-scope system". The Meeting Notetaker (launched 5 August 2026) "was not assessed in the SOC 2 Type I, the SOC 2 Type II, or the ISO 27001 certification audit." - **HIPAA:** "HIPAA-ready dictation requires a signed Business Associate Agreement (BAA)." Individual users can accept a BAA in the app under Settings → Data & Privacy; the [pricing page](https://wisprflow.ai/pricing) says "HIPAA-ready dictation with a signed BAA" under "Security and privacy. On every plan." and lists "Org-wide HIPAA enforcement with a signed BAA" on the Growth ($23 per user per month, $18 billed annually) and Enterprise plans. - **What a BAA changes:** according to Wispr's [HIPAA help article](https://docs.wisprflow.ai/articles/4218718535-hipaa-compliance-healthcare-use), signing it locks "Improve the model for everyone" and "Dictation Cloud Storage" off and blocks Notetaker recording. - **DPA:** a [Data Processing Addendum](https://wisprflow.ai/legal/dpa) (last updated 24 August 2026) incorporates the EU Standard Contractual Clauses (Commission Decision 2021/914, Module Two) and the UK Addendum, gives 10 days' notice of new sub-processors and deletes or returns personal data within 90 days after termination. Note the plan matrix on the pricing page: the "MSA & DPA" row is ticked only for Growth and Enterprise (Enterprise with custom terms), and "Models never trained on your data" only for Enterprise; on Free and Pro, training is an opt-out setting and you should not assume a signed DPA. - **Privacy policy:** the [privacy policy](https://wisprflow.ai/privacy-policy) (updated 19 August 2026) names no Standard Contractual Clauses, no EU-US Data Privacy Framework, no EU representative under Art. 27 GDPR and no data protection officer, and says content sent to AI providers "is generally deleted within 30 days". That sits uneasily with the zero-retention wording below; clarify it in your contract. - **Sub-processors:** "the Trust Center holds the current subprocessor list". The public page shows five logos (AWS, Baseten, Vercel, OpenAI, Supabase); independent analyses in 2026 counted more providers than that public summary. - **AI providers:** the [data-controls page](https://wisprflow.ai/data-controls) states: "Wispr always maintains zero data retention agreements with all third-party AI providers." **Where Wispr Flow is genuinely strong:** speed, polish and English accuracy, apps on all four major platforms, Command Mode, a free tier, and a self-serve BAA that small US practices can use without talking to sales. If you are a US clinic, that package is hard to beat. The [Sprecho vs Wispr Flow comparison](https://sprecho.ai/compare/sprecho-vs-wispr-flow) lists the full feature picture. ## What HIPAA, SOC 2 Type II and ISO 27001 actually attest **Each of the three documents answers one narrow question well.** The mistake is to read any of them as an answer to the German question of lawful disclosure. ### HIPAA and the Business Associate Agreement - **What it is:** the Health Insurance Portability and Accountability Act of 1996 and its Privacy and Security Rules, enforced by the US Department of Health and Human Services (HHS). - **What a BAA does:** it is the contract a US covered entity must sign before a vendor handles protected health information (PHI) on its behalf, obliging the vendor to safeguard PHI and report breaches. - **There is no official HIPAA certificate:** HHS does not certify vendors, so "HIPAA compliant" is always a vendor statement backed by a contract, not an audit seal. - **What it does not cover:** lawyers and tax advisors at all, German patients' rights under the GDPR, the third-country transfer question, and any duty under German criminal law. ### SOC 2 Type II - **What it is:** an attestation report by an independent CPA firm under the AICPA Trust Services Criteria (security, availability, processing integrity, confidentiality, privacy). - **Type I vs Type II:** Type I checks whether controls are designed properly at one point in time; Type II tests whether they actually operated over a period, typically several months. - **Scope is chosen by the company:** Wispr names dictation as the in-scope system, which is why the Notetaker sits outside it. - **Only as good as the auditor:** Wispr's earlier SOC 2 and ISO 27001 reports were issued via Delve, the compliance vendor accused in March 2026 of producing near-identical boilerplate reports (an investigation counted 494 reports with 99.8 % identical text). Wispr obtained a SOC 2 Type I from A-LIGN in April 2026 and now lists a Type II as current. ### ISO/IEC 27001:2022 - **What it is:** a certificate from an accredited certification body that the company runs an information security management system (ISMS) with risk treatment and Annex A controls. - **What it proves:** security is managed systematically and re-audited in a three-year cycle with annual surveillance audits. - **What it does not prove:** lawful processing under the GDPR, a lawful transfer to the US, or that a professional's secret is protected against foreign authorities. | Instrument | Who stands behind it | What it attests | What it does not answer in Germany | |---|---|---|---| | **HIPAA + BAA** | US law (HHS); contract between covered entity and vendor | Vendor contractually safeguards US protected health information | § 203 StGB, § 43e BRAO, § 62a StBerG, GDPR transfer rules; no protection for legal or tax secrets | | **SOC 2 Type II** | Independent CPA firm, AICPA criteria | Chosen controls operated effectively over a period | Where data may go, who may access it under foreign law, written secrecy undertaking | | **ISO/IEC 27001:2022** | Accredited certification body | An ISMS exists, is maintained and audited | CLOUD Act exposure, comparability of protection abroad (§ 43e Abs. 4 BRAO) | | **DPA with SCCs** | Contract under Art. 28 and Art. 46 GDPR | Processor duties and a GDPR transfer mechanism | The criminal-law secrecy undertaking required by § 203 Abs. 4 StGB and § 43e Abs. 3 BRAO | | **§ 203 confidentiality undertaking** | Contract in text form under German professional law | Provider is bound to secrecy, instructed about criminal consequences, limited to need-to-know | Security quality itself, which is why you still want the audits above | ## What German professional secrecy actually requires **German law asks three things a US certificate never asks:** is the provider allowed to learn the secret, is it bound to secrecy in a specific written form, and is the protection where the service is performed comparable to Germany? The wording below is quoted from the official texts of [§ 203 StGB](https://www.gesetze-im-internet.de/stgb/__203.html), [§ 43e BRAO](https://www.gesetze-im-internet.de/brao/__43e.html) and [§ 62a StBerG](https://www.gesetze-im-internet.de/stberg/__62a.html) on gesetze-im-internet.de. ### § 203 StGB: the criminal core - **Who is bound:** § 203 Abs. 1 Nr. 1 covers the "Arzt, Zahnarzt, Tierarzt, Apotheker" and other regulated health professions; Nr. 3 covers the "Rechtsanwalt, Kammerrechtsbeistand, Patentanwalt, Notar, Verteidiger […], Wirtschaftsprüfer, vereidigtem Buchprüfer, Steuerberater, Steuerbevollmächtigten". - **The offence:** whoever "unbefugt ein fremdes Geheimnis […] offenbart" faces "Freiheitsstrafe bis zu einem Jahr oder […] Geldstrafe"; up to two years if done for payment or to cause harm (Abs. 6). - **The service-provider gateway (Abs. 3 Satz 2):** professionals "dürfen fremde Geheimnisse gegenüber sonstigen Personen offenbaren, die an ihrer beruflichen oder dienstlichen Tätigkeit mitwirken, soweit dies für die Inanspruchnahme der Tätigkeit der sonstigen mitwirkenden Personen erforderlich ist". A dictation provider can be such a contributing person. - **Your own liability (Abs. 4 Satz 2 Nr. 1):** you are punishable if you have "nicht dafür Sorge getragen" that the contributing person "zur Geheimhaltung verpflichtet wurde". In plain English: if the provider leaks and you never bound it to secrecy, you are liable too. ### § 43e BRAO and § 62a StBerG: the contract rules - **Careful selection (Abs. 2):** the lawyer "ist verpflichtet, den Dienstleister sorgfältig auszuwählen" and must end the cooperation without delay if compliance is not ensured. - **Text form (Abs. 3):** "Der Vertrag mit dem Dienstleister bedarf der Textform." It must (1) bind the provider to secrecy "unter Belehrung über die strafrechtlichen Folgen einer Pflichtverletzung", (2) limit knowledge of secrets to what the contract requires, and (3) state whether subcontractors may be used, who must in turn be bound in text form. - **Services abroad (Abs. 4):** access to secrets is allowed only "wenn der dort bestehende Schutz der Geheimnisse dem Schutz im Inland vergleichbar ist". - **Data protection stays separate (Abs. 8):** "Die Vorschriften zum Schutz personenbezogener Daten bleiben unberührt." You need the AVV and the secrecy undertaking. - **Tax advisors:** § 62a StBerG repeats the same structure almost word for word for Steuerberater and Steuerbevollmächtigte, including Abs. 4: services abroad only where secrecy protection there is comparable to Germany. - **Doctors:** there is no § 43e equivalent in a federal statute, but § 203 Abs. 3 and 4 StGB apply in full, together with the professional code of the regional medical chamber. ### GDPR Art. 9 and 28, the CLOUD Act and Schrems II - **Art. 9 GDPR:** health data is a special category. Every Arztbrief, Befund or therapy note dictated through a cloud tool is special-category data in transit. - **Art. 28 GDPR:** a processor needs a written contract (AVV), sufficient guarantees and authorised sub-processors you can actually see. - **Chapter V GDPR:** a transfer to the US needs a legal basis such as the SCCs plus a transfer impact assessment; the CJEU's Schrems II judgment (C-311/18, 16 July 2020) made that assessment mandatory in practice. - **CLOUD Act (2018):** US providers can be compelled to hand over data in their "possession, custody, or control" regardless of where the server stands, which is the core of the comparability question in § 43e Abs. 4 BRAO. ## Applying the law to Wispr Flow's published facts **Measured against its own documentation, Wispr Flow can satisfy the GDPR paperwork but leaves the § 203 questions with you.** We do not claim Wispr Flow is unlawful for every professional; we show where each test lands, using only what Wispr publishes. ### ![Wispr Flow logo](https://sprecho.ai/images/competitors/wispr-flow.svg) Wispr Flow against the five German tests 1. **Text-form secrecy undertaking:** Wispr's DPA is a GDPR addendum under US contract terms. A § 43e-style undertaking with the instruction about criminal consequences is a separate German document; if you need it, you have to request it, negotiate it and verify it yourself. 2. **Need-to-know limitation:** the BAA locks storage and model training off for dictation, which helps. Context Awareness, by Wispr's wording, can use "relevant text data from the active app window", so in a case file that context may itself be a secret. Keep it off. 3. **Sub-processors bound in text form:** the list sits behind the Trust Center and can change with 10 days' notice. You must check that every provider in the chain, including the AI providers covered by zero-retention agreements, is bound as § 43e Abs. 3 Nr. 3 demands. 4. **Comparable protection abroad:** processing happens in the United States, by a US company subject to the CLOUD Act. That is exactly the case § 43e Abs. 4 BRAO and § 62a Abs. 4 StBerG single out, and many chambers and DPOs treat it as the hardest hurdle. 5. **GDPR transfer basis:** the DPA (Growth and Enterprise) uses SCCs Module Two, but the transfer impact assessment is your job as controller. Neither the DPA nor the privacy policy names an EU representative under Art. 27 GDPR. **Practical verdict:** a German professional can use Wispr Flow for content that contains no client or patient secrets (marketing copy, general research, internal notes without names). For mandate files, Befunde and tax matters, the safer reading of § 203 and § 43e is that a US-processed service is hard to justify, regardless of HIPAA. ### ![Sprecho logo](https://sprecho.ai/sprecho-icon.svg) Sprecho against the same tests **Sprecho is a voice dictation app by Melo Designer GmbH (Lastrup, Germany) that writes formatted text into any app on Windows, macOS, Linux, iOS and Android, with all processing in the EU.** Against the same five tests: - **Secrecy undertaking:** for Enterprise customers, Sprecho signs a separate confidentiality agreement under § 203 Abs. 4 StGB, § 43e Abs. 3 BRAO or § 62a StBerG, in addition to the AVV, because an AVV alone does not cover the criminal-law instruction. - **Need-to-know:** no third-party AI provider; speech recognition and formatting run on models Sprecho operates. Customer audio and transcripts are never used for training. - **Sub-processors:** a public list (STRATO GmbH for hosting, Media Trooper GmbH for GPU inference in Germany and the Netherlands, Hetzner in Nuremberg for off-site backups only) with 30 days' notice of material changes. See [where Sprecho runs](https://sprecho.ai/help/privacy-and-security/where-sprecho-runs-and-who-processes-your-data). - **Comparable protection:** German company, EU/EEA processing only, no US entity in the dictation path, so no third-country transfer to assess. - **GDPR:** a public AVV under Art. 28 for every business customer, countersigned within 5 working days. Details in [GDPR and your data processing agreement](https://sprecho.ai/help/privacy-and-security/gdpr-and-your-data-processing-agreement). ## Profession by profession: duty, Wispr Flow, Sprecho **The same tool lands differently for each profession**, because the duties come from different statutes. The table summarises the published facts; it is not legal advice for your specific case. | Profession | Core duty | What Wispr Flow offers | What Sprecho offers | |---|---|---|---| | **Rechtsanwalt / law firm** | § 203 Abs. 1 Nr. 3 StGB, § 43a Abs. 2 and § 43e BRAO: text-form undertaking, comparable protection abroad | US processing, DPA with SCCs, BAA (health data only), Trust-Center sub-processors; US processing ❌ | EU processing, public AVV, § 203 / § 43e confidentiality agreement (Enterprise) ✅ | | **Arzt / Psychotherapeut** | § 203 Abs. 1 Nr. 1 StGB, Art. 9 GDPR (health data), professional code of the medical chamber | Self-serve BAA locks storage and training off; US processing ❌ | EU processing, delete-immediately retention, AVV and confidentiality agreement (Enterprise) ✅ | | **Steuerberater / tax firm** | § 203 Abs. 1 Nr. 3 StGB, § 57 and § 62a StBerG | Same as for lawyers; HIPAA irrelevant to tax secrets; US processing ❌ | EU processing, works in DATEV, Word and Outlook, confidentiality agreement under § 62a StBerG (Enterprise) ✅ | | **Staff and trainees** | § 203 Abs. 3 Satz 1 and Abs. 4 StGB (Gehilfen) | SSO/SAML and admin control over model training from Growth ($23 per user per month); SCIM, MDM and audit logs on Enterprise only | Enterprise: SSO/SAML, MDM, audit-log export, shared team dictionary | ## How a compliant setup looks in practice **A compliant dictation setup is mostly configuration and paperwork, done once.** These are the setups we recommend to German professionals, built on the features documented in Sprecho's help center. ### Kanzlei: briefs, letters and file notes - **Dictate where you write:** push-to-talk hotkey into Word, Outlook, the browser or case-management software such as RA-MICRO; Sprecho also works in Citrix and RDP sessions. - **Snippets for boilerplate:** trigger phrases for recurring Schriftsatz openings, evidence offers and signatures, with {date} and {time} variables. - **Dictionary:** case numbers, party names and statute references with the right capitalisation; see [teach Sprecho your words](https://sprecho.ai/help/formatting-and-ai/teach-sprecho-your-words-with-the-dictionary). - **§ sign by voice:** "section sign" (German: "Paragraphenzeichen") is a documented voice command, as are new line and new paragraph. - **Paperwork:** AVV plus confidentiality agreement (Enterprise), filed with your § 43e documentation. Our guide to [GDPR speech recognition for law firms](https://sprecho.ai/blog/gdpr-speech-recognition-for-lawyers-law-firms) goes deeper. ### Arztpraxis: Arztbriefe and Befunde ![Doctor at a German practice reception dictating into a smartphone, where patient data falls under Art. 9 GDPR and § 203 StGB](https://api.sprecho.ai/media/2026/10/1791204816-d7a77977.jpg) - **Retention first:** set text and audio retention to "Delete immediately after pasting" under Settings → Privacy → Data Retention. The defaults are 14 days for audio and never for text, so this step matters; it is enforced server-side, as described in [turn off transcript and audio storage](https://sprecho.ai/help/privacy-and-security/turn-off-transcript-and-audio-storage). - **Practice software:** dictate into the PVS, the browser or Word; the medical vocabulary lives in a shared practice dictionary. - **Snippets for findings:** normal-findings text, follow-up phrases and signature blocks by trigger phrase. - **Mobile:** on iPhone and Android, Sprecho dictates through its own keyboard (on Android also via the floating bubble), with the same account and dictionary across up to 3 devices. - **More detail:** [Sprecho for doctors](https://sprecho.ai/solutions/healthcare) and our [Arztbriefe guide](https://sprecho.ai/blog/gdpr-speech-recognition-for-doctors-arztbriefe). ### Steuerkanzlei: Einsprüche and client letters ![Tax advisor dictating client correspondence at a desk with unlabelled folders in a German tax office](https://api.sprecho.ai/media/2026/10/1791204817-113cc3ff.jpg) - **DATEV and Outlook:** dictate Einspruch reasoning and client mails directly into the application you already use. - **Snippets:** standard appeal wording and power-of-attorney text by trigger phrase, shared across the team. - **Writing style per app:** formal for letters to the tax office, factual for client mail. - **Paperwork:** AVV plus confidentiality agreement under § 62a StBerG (Enterprise). See [Sprecho for tax consultants](https://sprecho.ai/solutions/tax-consultant) and the [DATEV dictation guide](https://sprecho.ai/blog/gdpr-speech-recognition-for-tax-advisors-datev). ## The data protection officer's checklist **Use this checklist before any dictation tool touches mandate or patient data.** The Wispr Flow column reflects what Wispr publishes as of 5 October 2026; anything marked "on request" needs to be obtained and read, not assumed. | Check | Legal basis | Wispr Flow (as published) | Sprecho | |---|---|---|---| | Provider seat and jurisdiction | § 43e Abs. 4 BRAO, CLOUD Act | ❌ Wispr AI, Inc., Delaware, US | ✅ Melo Designer GmbH, Germany | | Processing location | Chapter V GDPR | ❌ United States | ✅ Germany and the Netherlands | | AVV / DPA | Art. 28 GDPR | ⚠️ DPA with SCCs Module Two; "MSA & DPA" ticked for Growth and Enterprise only | ✅ Public AVV, German and English | | Text-form secrecy undertaking with criminal-law instruction | § 203 Abs. 4 StGB, § 43e Abs. 3 BRAO, § 62a StBerG | ⚠️ Not part of the published DPA; negotiate separately | ✅ Separate confidentiality agreement for Enterprise customers | | Sub-processor list you can read | Art. 28 Abs. 2 GDPR, § 43e Abs. 3 Nr. 3 BRAO | ⚠️ Behind Trust-Center access; 10 days' notice of changes | ✅ Public list; 30 days' notice | | Third-party AI providers | Need-to-know, Art. 5 GDPR | ⚠️ Yes; zero-retention agreements per data-controls page, "generally deleted within 30 days" per privacy policy | ✅ None | | Retention you control | Art. 5 Abs. 1 lit. e GDPR | ✅ Dictation Cloud Storage off; locked off with BAA | ✅ Delete immediately after pasting, server-side | | Transfer impact assessment | Schrems II, C-311/18 | ❌ Required, your responsibility | ✅ Not needed (no third-country transfer) | | Security attestations | Art. 32 GDPR | ✅ SOC 2 Type II, ISO/IEC 27001:2022 | ✅ TOMs in the AVV; DSB TÜV SÜD-certified | | Price per seat (net) | Budget | $15 Pro / $23 Growth per month (BAA org-wide from Growth) | €10.92 Pro / €21.84 Enterprise per month; €9.24 / €18.47 billed yearly | **A note on prices:** Wispr's US pricing page shows USD; German visitors have seen the same digits in euros (€15 / €12 for Pro, observed in May 2026). Wispr AI, Inc. invoices from the US and, per its [help article on VAT](https://docs.wisprflow.ai/articles/6671605687), "we do not add VAT, GST, or other local sales tax". Business buyers therefore account for German VAT themselves via reverse charge; for medical practices whose services are VAT-exempt and which cannot deduct input tax, that 19 % is a real extra cost. Sprecho's prices are net plus 19 % VAT on a German invoice; see [pricing](https://sprecho.ai/pricing). ## When Wispr Flow is a reasonable choice **Wispr Flow is a reasonable choice whenever no professional secret enters the microphone.** Being precise about this is part of being fair. - **English-first work without client data:** marketing texts, academic writing, general correspondence. - **US healthcare organisations:** exactly the audience the BAA was built for. - **Teams that need meeting notes:** the Notetaker has no Sprecho equivalent, but remember it sits outside the audit scope and is blocked once a BAA is signed. - **Not a fit:** mandate files, patient documentation, tax files, and public-sector work in Germany, where most DPOs require EU processing. ## Conclusion HIPAA, SOC 2 Type II and ISO 27001 show that Wispr Flow takes security seriously, and for US clinics they are the right documents. For German lawyers, doctors and tax advisors the decisive questions are different: a text-form secrecy undertaking, a visible and bound sub-processor chain, and protection abroad that is comparable to Germany. Wispr Flow's US processing leaves those questions open and puts the burden on you. **If you dictate client or patient secrets, start Sprecho's 14-day trial without a card, sign the AVV and, on Enterprise, request the confidentiality agreement**, and compare the details on our [Sprecho vs Wispr Flow page](https://sprecho.ai/compare/sprecho-vs-wispr-flow) or in the overview of [Wispr Flow alternatives for Germany](https://sprecho.ai/blog/wispr-flow-alternatives-germany). ## FAQ ### Is Wispr Flow HIPAA compliant? Wispr says yes, provided a Business Associate Agreement is signed: individual users can accept a BAA in the app, and Growth and Enterprise plans enforce it organisation-wide. Signing locks Dictation Cloud Storage and model training off and blocks Notetaker recording. HIPAA is a US law, so this tells a German practice little about § 203 StGB. ### Is HIPAA relevant for doctors in Germany? Not legally. HIPAA applies to US covered entities and their business associates; German doctors are bound by § 203 Abs. 1 Nr. 1 StGB, Art. 9 GDPR and the professional code of their medical chamber. A BAA can be useful evidence of security practice, but it neither replaces an AVV nor answers the third-country question. ### Can a German lawyer use Wispr Flow for client matters? Only if the firm can satisfy § 43e BRAO: a text-form contract binding the provider to secrecy with an instruction about criminal consequences, bound subcontractors, and protection abroad comparable to Germany (§ 43e Abs. 4). Because Wispr Flow processes data in the US under US jurisdiction, most DPOs advise against using it for mandate content. Content without client secrets is a different matter. ### Can a doctor dictate patient letters with Wispr Flow under Art. 9 GDPR? Health data is a special category under Art. 9 GDPR and is additionally protected by § 203 StGB. Using Wispr Flow would mean a US transfer based on SCCs plus a transfer impact assessment, a BAA to stop storage, and a separate secrecy undertaking. It is possible on paper but hard to defend, which is why EU-processed tools are the usual recommendation. ### What is § 43e BRAO? § 43e of the Federal Lawyers' Act (BRAO) sets the rules for lawyers using external service providers, including IT and cloud services. It requires careful selection, a contract in text form with a secrecy obligation and an instruction about criminal consequences, rules for subcontractors, and comparable protection for services abroad. § 62a StBerG mirrors it for tax advisors. ### Does Wispr Flow sign an AVV (DPA)? Wispr Flow offers a Data Processing Addendum that incorporates the EU Standard Contractual Clauses (Module Two) and the UK Addendum, last updated on 24 August 2026; the pricing table ticks "MSA & DPA" only for Growth and Enterprise. That covers Art. 28 GDPR formally, but it is a GDPR document; the criminal-law secrecy undertaking required by § 203 Abs. 4 StGB and § 43e Abs. 3 BRAO is a separate matter. ### What does Sprecho provide for § 203 StGB? Sprecho, from Melo Designer GmbH in Germany, processes all dictation in the EU with no third-party AI provider and publishes its AVV and sub-processor list. For Enterprise customers it signs a separate confidentiality agreement under § 203 Abs. 4 StGB, § 43e BRAO or § 62a StBerG. Retention can be set to delete immediately after pasting, enforced on the server. ### Is my voice biometric data under the GDPR? Voice becomes biometric data under Art. 4 Nr. 14 and Art. 9 GDPR only when it is processed to uniquely identify a person, which plain dictation does not do. The content you dictate is usually the bigger issue: patient details are health data under Art. 9, and client facts are secrets under § 203 StGB regardless of their GDPR category. --- ## https://sprecho.ai/blog/switch-from-wispr-flow-to-sprecho-in-20-minutes — Switching from Wispr Flow to Sprecho in 20 Minutes: Dictionary, Snippets, Hotkey, Styles > Switch from Wispr Flow to Sprecho in 20 minutes: rebuild dictionary and snippets, set the hotkey, pick per-app styles, turn storage off. Honest pros and cons. *Published: 2026-11-20T18:03:33.591000Z* *Tags: wispr-flow, migration, alternatives, voice-dictation, gdpr, germany, comparison* To **switch from Wispr Flow to Sprecho**, you do not need an export tool, a consultant or a free afternoon. You need about 20 minutes: install Sprecho, set your dictation key, re-enter the dictionary words and snippets you actually use, pick a formatting model per app, and switch transcript storage off. This guide walks through each step with the exact labels from the Sprecho app, and says plainly what you leave behind. *Last verified: October 2026.* - **Time needed:** about 20 minutes for one person with roughly 50 dictionary words and 10 snippets; teams add a short rollout on top. - **What moves automatically:** nothing. There is no Wispr Flow import in Sprecho, and Wispr's help center documents no dictionary export, so you re-enter entries by hand. - **What you gain:** processing in Germany and the Netherlands by a German GmbH, a public AVV, a Linux app, a German invoice and formatting that works in German. - **What you lose:** Wispr's Command Mode, Flow Notetaker and Wispr's style presets. ## Why do German users switch from Wispr Flow to Sprecho? Most people who switch do so for **jurisdiction, language and billing**, not because Wispr Flow is a bad product. Wispr Flow is an AI voice dictation app made by Wispr AI, Inc., a Delaware corporation based in San Francisco, that turns speech into polished text in any app on Mac, Windows, iPhone and Android. Sprecho is a voice dictation app made by Melo Designer GmbH in Lastrup, Germany, that does the same job on Windows, macOS, Linux, iOS and Android, with all processing in the EU. - **US data processing:** Wispr's own [security FAQ](https://docs.wisprflow.ai/articles/3467817258-security-and-compliance-faq) states that "Wispr Flow is a cloud service … processes and stores customer data in the United States." For a German company, that is a third-country transfer that needs a transfer impact assessment. - **CLOUD Act exposure:** a US company can be compelled under the CLOUD Act (2018) to hand over data it controls, wherever the server stands. - **Styles in English only:** Wispr's [help article on Flow Styles](https://docs.wisprflow.ai/articles/2368263928-how-to-setup-flow-styles) states that "style formatting applies in English only". German dictations do not get the Formal, Casual or Excited treatment. - **Professional secrecy:** lawyers, doctors and tax advisors bound by **§ 203 StGB** need a processor they can bind by contract under German law; see our [solution for law firms](/solutions/legal) and [for practices and clinics](/solutions/healthcare). - **Invoicing:** Wispr AI, Inc. bills from the US and adds no VAT (German businesses book reverse charge); Sprecho invoices as a German GmbH with a VAT line. To be fair: Wispr Flow is fast, polished and well funded. It raised a $280 million Series B at a $2 billion valuation, announced on 17 August 2026 according to [TechCrunch](https://techcrunch.com/2026/08/17/wispr-raises-280m-at-2b-valuation-as-it-looks-beyond-dictation/), and ships features quickly. If you dictate mostly in English and have no secrecy duties, it is a strong product. The full comparison is on our [Sprecho vs Wispr Flow page](/compare/sprecho-vs-wispr-flow), and the wider market is covered in [the best Wispr Flow alternatives for Germany](/blog/wispr-flow-alternatives-germany). ## Which Wispr Flow features have a Sprecho equivalent? Almost every everyday Wispr Flow feature has a **direct Sprecho counterpart**; the gaps are Command Mode and the meeting notetaker. The table maps each feature to its Sprecho name and the place you find it. | Wispr Flow feature | Sprecho equivalent | Where in Sprecho | Match | |---|---|---|---| | Push-to-talk (Fn on Mac, Ctrl+Win on Windows) | Hold-to-talk on the dictation key | fn (macOS), Insert (Windows), Control+Space (Linux) | ✅ Same idea | | Hands-free (Fn+Space, Ctrl+Win+Space) | Lock mode: tap the same key once | No setting needed | ✅ Same idea | | Dictionary with "Correct a misspelling" | Dictionary with **Correct a misspelling** | Sidebar → Dictionary → Add new | ✅ Re-enter by hand | | Dictionary auto-learning | Auto-learn plus a Suggestions inbox | Dictionary page | ✅ | | Snippets (trigger → text) | Snippets with **Trigger phrase** and **Expands to** | Sidebar → Snippets → Add new | ✅ Plus {date}, {time}, {clipboard} | | Team dictionary and snippets | **Share with team** toggle | Dictionary and Snippets, on Pro and Enterprise | ✅ | | Flow Styles per app (English only) | Formatting model per app category | Sidebar → Models | ✅ Works in German | | Filler removal, self-correction | Automatic punctuation and filler removal; "scratch that" | Always on | ✅ | | Command Mode (voice edits on text) | No equivalent | n/a | ❌ | | Flow Notetaker (meetings) | No meeting recorder; audio-file import and Notes instead | Notes, audio import | ❌ Different scope | | Dictation Cloud Storage setting | **Text Retention** and **Audio Retention** | Settings → Privacy → Data Retention | ✅ Enforced server-side | | Linux app | Linux desktop app | Download page | ✅ Sprecho only | - **Read the "Match" column first:** every ✅ row is a setting you reproduce in the next 20 minutes; every ❌ row is a habit you replace. - **Keyboard muscle memory:** the hold-versus-tap logic is the same, only the default key differs, and you can rebind it. ## What transfers from Wispr Flow, and what has to be rebuilt? **Nothing transfers automatically**: Sprecho has no Wispr Flow import, and Wispr's [dictionary article](https://docs.wisprflow.ai/articles/4052411709-teach-flow-your-words-with-the-dictionary) documents no export or CSV option. Wispr Flow syncs your dictionary and snippets between its own apps, but there is no file you can take with you. The good news is that only a few things are worth moving, and most of them take minutes. | What | Transfers automatically? | How to move it | Typical time | |---|---|---|---| | Dictionary words | ❌ No | Copy from Wispr's Dictionary screen, add in Sprecho with **Add new** | 5–6 minutes for ~50 words | | Misspelling rules | ❌ No | Re-create with **Correct a misspelling** | 1–2 minutes | | Snippets | ❌ No | Re-create with **Trigger phrase** and **Expands to** | 3–4 minutes for ~10 snippets | | Style choices | ❌ No | Pick Standard or Thinking per tab on the **Models** page | 2 minutes | | Hotkey habit | ❌ No | Keep Sprecho's default or rebind under **Settings → General** | 2 minutes | | Dictation history | ❌ No | Leave it in Wispr or delete it there; start fresh | 0 minutes | | Notetaker transcripts | ❌ No | Download what you need from Wispr before cancelling | Depends on volume | | Subscription | ❌ No | Cancel with Wispr's billing provider; see below | 2 minutes | - **Prune while you copy:** a dictionary built over months contains words you added once and never needed again. Re-entering is a chance to keep only what you use. - **Start with your 20 most-corrected names:** they deliver most of the accuracy gain on day one; auto-learning picks up the rest. - **Do not copy history:** starting with an empty history is also the cleanest privacy position. ## How do you switch in 20 minutes, step by step? Work through **seven steps in order**; the times are our estimate for one person, measured from opening the download page. Keep Wispr Flow installed until step 7 so you can read its Dictionary and Snippets screens while you rebuild them. | Step | Minutes | What you do | Done | |---|---|---|---| | 1. Install and sign in | 0–3 | Install Sprecho, create an account, start the 14-day Pro trial (no card) | ☐ | | 2. Set the dictation key | 3–5 | Keep fn / Insert / Control+Space or rebind under **Settings → General** | ☐ | | 3. Rebuild the dictionary | 5–11 | Names, product terms, misspelling rules; pin the critical ones | ☐ | | 4. Rebuild snippets | 11–15 | Signature, email, address, boilerplate; add {date} where useful | ☐ | | 5. Choose formatting per app | 15–17 | **Models** page: Standard or Thinking per category | ☐ | | 6. Switch storage off | 17–19 | **Text Retention** → Delete immediately after pasting | ☐ | | 7. Test and cancel Wispr Flow | 19–20 | Dictate one email in German, then cancel Wispr's subscription | ☐ | ### Step 1: install Sprecho and start the trial (minutes 0–3) Install the desktop app and sign in; the **14-day Pro trial starts without a credit card**. - **Windows 10 or 11 (64-bit):** download the Windows installer; IT departments get an **.msi** build for Intune or Group Policy. Details in [Install Sprecho on Windows](/help/getting-started/install-sprecho-on-windows). - **macOS:** Apple Silicon and Intel are supported; grant the microphone and Accessibility permissions when asked. - **Linux:** a .deb package for Debian and Ubuntu is available, something Wispr Flow does not offer. - **Phone:** install the Sprecho app from the App Store or Google Play and sign in with the same account; your dictionary and snippets follow you. One licence covers **three devices**. ### Step 2: set the dictation key (minutes 3–5) Sprecho uses **one key for both modes**: hold it for push-to-talk, tap it once for hands-free lock mode. Wispr Flow uses separate shortcuts for the two, so this is the one habit that changes. - **Defaults:** `fn` on macOS, `Insert` on Windows, `Control`+`Space` on Linux. - **Coming from Wispr on a Mac:** Wispr's push-to-talk default is also `Fn`, so your thumb already knows the key. Quit Wispr Flow before testing so the two apps do not compete for it. - **Rebind:** open **Settings → General**, click the **Dictation Hotkey** field under **Keyboard Shortcuts** and press the combination you want. It saves immediately. - **Second key:** a **secondary shortcut** can trigger the same dictation, useful for an external keyboard. - **Cancel:** `Esc` discards a running recording; nothing is transcribed or sent. The full key list is in the [keyboard shortcuts cheat sheet](/help/getting-started/keyboard-shortcuts-cheat-sheet) and [hold-to-talk vs lock mode](/help/dictation/hold-to-talk-vs-lock-mode). ### Step 3: rebuild your dictionary (minutes 5–11) Open Wispr Flow's Dictionary screen in one window and **Sprecho's Dictionary** in the other, then re-enter the entries you still need. ![Notebook with a hand-drawn checklist being ticked off during a 20-minute switch from Wispr Flow to Sprecho](https://api.sprecho.ai/media/2026/10/1791204770-53822a00.jpg) 1. **Open the page:** click **Dictionary** in the Sprecho sidebar, then **Add new**. 2. **Add plain words:** type each term under **Add a new word** exactly as it should appear, capitalisation included, and click **Add word**. 3. **Re-create replacement rules:** switch on **Correct a misspelling** and fill in **Misspelling** and **Correct spelling**, for example "schmitt" → "Schmidt". 4. **Pin what must never fail:** the **Pin to recognizer** action reserves a slot for your own name, your biggest client or the statute you cite daily. Keep the pinned list short. 5. **Share company terms:** on a team, switch on **Share with team** so colleagues do not add the same word again. - **Auto-learning:** when you edit a dictation in your history, clear corrections are learned straight away with an **Undo** button, and less certain ones land in **Suggestions**. - **Budget:** the recogniser receives the 200 most relevant entries per dictation, pinned entries first, so a lean list works better than a long one. More in [Teach Sprecho your words with the dictionary](/help/formatting-and-ai/teach-sprecho-your-words-with-the-dictionary). ### Step 4: rebuild your snippets (minutes 11–15) Snippets work the same way as in Wispr Flow: you say a **trigger phrase** and Sprecho inserts the stored text. 1. **Open the page:** click **Snippets** in the sidebar, then **Add new**. 2. **Trigger phrase:** what you will say; matching is case-insensitive and tolerant for triggers of five characters or more. 3. **Expands to:** the text to insert; this is a rich editor with paragraphs, headings, quotes and links. 4. **Save:** click **Add snippet**. - **Dynamic variables:** `{date}`, `{time}` and `{clipboard}` are filled in at the moment the snippet expands, which turns a meeting-note snippet into a dated template. - **Formatting follows the app:** Slack and WhatsApp receive markdown, Outlook receives rich text, so one signature works everywhere. - **Trigger hygiene:** "my postal address" is safe, "address" fires mid-sentence. Two or three words that only occur together are ideal. Details in [Snippets: voice shortcuts for text you repeat](/help/formatting-and-ai/snippets-voice-shortcuts-for-text-you-repeat). ### Step 5: choose formatting per app (minutes 15–17) Instead of Wispr's tone presets, Sprecho lets you choose **how much editing** each app category gets, and that choice works for German dictations as well as English ones. - **Where:** click **Models** in the sidebar and pick a tab: **Personal messages**, **Work messages**, **Email**, **Code** or **Other**. - **Standard:** keeps your wording exactly as spoken and cleans up punctuation, layout and lists. - **Thinking:** rewrites sentences for grammar and readability without changing the meaning; it is the shipped default for Email. - **Mapping from Wispr:** Wispr's Personal, Work, Email and Other categories map one to one; Sprecho adds a separate **Code** category so identifiers are never rewritten. - **Per device:** the setting applies to every dictation from that computer, on macOS and Windows. Sprecho has **no Formal, Casual or Excited switch**; the models take their cue from the text itself. If you relied on "Excited!" in Wispr, that preset has no counterpart. See [Set a writing style per app](/help/formatting-and-ai/set-a-writing-style-per-app). ### Step 6: switch transcript storage off (minutes 17–19) Set Sprecho to **keep nothing** before you dictate anything confidential. 1. Open **Settings → Privacy** and scroll to **Data Retention**. 2. Set **Text Retention** to **Delete immediately after pasting**. 3. Set **Audio Retention** to **Delete immediately after pasting**. - **Enforced on the server:** with text set to delete immediately, the transcript row is never written, so nothing can sync to other devices. - **Defaults otherwise:** 14 days for audio and never delete for text, with fixed windows from 1 day to 3 months in between. - **Trade-off:** you lose the history list and re-copying old dictations; dictionary, snippets and formatting keep working. The mechanism is explained in [Turn off transcript and audio storage](/help/privacy-and-security/turn-off-transcript-and-audio-storage). ### Step 7: test with a real email, then cancel (minutes 19–20) Dictate **one real German email** with the names from your dictionary and one snippet in it. If the names come out right and the snippet expands, the switch is done; cancel Wispr Flow as described below. ## How do voice commands work after the switch? Sprecho handles **punctuation and layout commands** in German, English and Portuguese, and adds punctuation on its own, so most people say fewer commands than in Wispr Flow. - **Layout:** "new line" and "new paragraph"; "paragraph" alone stays a word. - **Punctuation:** period, comma, colon, semicolon, question mark, exclamation mark, open/close quote, open/close bracket, section sign (§), slash, hyphen, underscore. - **Correction:** "scratch that" (German: "Streich das") removes what you just said. - **The pause rule:** leave about half a second before a command, otherwise it is written as words. - **Caps Lock:** hold Caps Lock while dictating and the output comes out in UPPERCASE, handy for headings and section markers. If you also come from Dragon, our [Dragon voice commands cheat sheet](/blog/dragon-voice-commands-in-sprecho-2026) maps the older command set. ## What do you lose when you leave Wispr Flow? You lose **three real features** and one speed claim. Be honest with yourself about how often you used them. - **Command Mode:** in Wispr Flow you hold a shortcut and speak an instruction such as editing, drafting a reply or searching; Wispr's [Command Mode article](https://docs.wisprflow.ai/articles/4816967992-how-to-use-command-mode) lists it as a paid-plan feature. Sprecho has no voice editing of existing text. - **Hands-free lock:** Wispr's docs state that saying "Hey Flow" during ordinary dictation does not start Command Mode; its hands-free lock is a double press of the shortcut. Sprecho always starts from the dictation key. - **Flow Notetaker:** Wispr's meeting recorder (Mac since 5 August 2026, Windows since 15 September 2026) has no Sprecho counterpart; Sprecho is a dictation tool and offers audio-file import instead. - **Style presets:** Formal, Casual, very casual and Excited! do not exist in Sprecho; for English users who liked them, that is a loss. - **English speed:** Wispr advertises very low English latency and its own Canto model, benchmarked on English recordings only. ## What do you gain with Sprecho? You gain **a German legal counterparty and German-first formatting**, which matters most if you dictate client, patient or employee data. - **EU processing:** servers at STRATO GmbH in Germany, GPU inference by Media Trooper in Germany and the Netherlands; no US sub-processor in the dictation path. - **Public AVV:** a data processing agreement under Art. 28 GDPR for every business customer, downloadable without a sales call; see [GDPR and your data processing agreement](/help/privacy-and-security/gdpr-and-your-data-processing-agreement). - **German formatting:** the formatting models work in every supported language, so German emails get the same cleanup as English ones. - **Linux:** a native Linux app next to Windows and macOS. - **Citrix and RDP:** dictation works in remote desktop and terminal-server sessions. - **German invoice:** prices in euros including 19 % VAT, invoiced by a GmbH. One caveat applies to both products: **neither works offline**. Sprecho is cloud-based too, it just runs in the EU. Background on where audio goes is in [where is my voice stored when I dictate with AI](/blog/where-is-my-voice-stored-ai-dictation). ## How does a team switch together? A team switches in **one shared setup plus 20 minutes per person**, and most of the dictionary work is done once for everyone. ![Team lead onboarding colleagues onto Sprecho at a meeting table with laptops](https://api.sprecho.ai/media/2026/10/1791204770-cfee3166.jpg) 1. **Create the team:** click **Team** in the sidebar, then **Add your team**; invite colleagues by email or with **Import CSV/Excel**. 2. **Build the shared dictionary first:** one person adds company names, product terms and abbreviations with **Share with team** switched on. 3. **Add shared snippets:** the standard reply, the disclaimer, the address block, so everyone uses one wording. 4. **Set a retention policy:** the team owner sets it once; members then see **Managed by your team** and the stricter setting applies. 5. **Sign the AVV:** download the PDF, sign it and send it to dpo@sprecho.ai; the countersigned copy comes back within five working days. 6. **Enterprise only if you need control:** admin dashboard, licence management, SSO/SAML, MDM deployment via Intune or Jamf, and pay by invoice. - **Sprecho Pro:** €12.99 per month or €10.99 per month billed yearly, both brutto including 19 % VAT (net €10.92 and €9.24); shared team dictionary and snippets are included. - **Sprecho Enterprise:** €25.99 or €21.99 per seat per month brutto (net €21.84 and €18.47), see [pricing](/pricing). - **Wispr Flow for comparison:** Pro at $15 per user per month or $12 billed yearly, Growth at $23 or $18; German visitors see the same digits in euros, and Wispr adds no VAT. - **Regulated teams:** see our pages for [tax advisors](/solutions/tax-consultant) and [public administration](/solutions/government). ## What happens to your Wispr Flow subscription and data? You **cancel with whoever bills you**, and your paid access runs until the end of the period. Wispr's [cancellation article](https://docs.wisprflow.ai/articles/3307336566-canceling-your-wispr-flow-subscription) is specific about this. - **Desktop or web purchase:** cancel through Wispr's Stripe billing. - **iPhone purchase:** cancel under iPhone Settings → your name → Subscriptions; Apple handles App Store refunds. - **Android purchase:** cancel in the Play Store under Payments & subscriptions. - **Access after cancelling:** paid features continue to the end of the billing period, then the account returns to the free tier; the account and data are not deleted. - **Refunds:** Wispr's article says there is "no automatic prorated refund" and asks you to contact support; check their terms for your case. - **Store subscriptions:** deleting your Wispr account or signing out does not cancel an App Store or Google Play subscription. - **Stored data:** according to Wispr's [data-controls page](https://wisprflow.ai/data-controls), the Dictation Cloud Storage setting "controls whether Wispr stores your transcription data (transcripts, audio, and dictation history) on Wispr's servers"; review it, or delete your account, if you want nothing left behind. ## Who should stay on Wispr Flow? Stay if your work is **English-first, unregulated and command-heavy**; switch if any of the following is false. - **Stay:** you dictate mainly in English, use Command Mode daily, rely on Notetaker for meetings and have no secrecy duties. - **Switch:** you dictate in German, handle client or patient data under § 203 StGB or Art. 9 GDPR, need an AVV with a German company, or use Linux. - **Test both:** both offer a free tier of 2,000 words per week on desktop, so you can run them side by side for a week. ## Conclusion Switching from Wispr Flow to Sprecho is a 20-minute job because the concepts are the same: a hold-or-tap dictation key, a dictionary with misspelling rules, snippets with trigger phrases and formatting per app. Nothing imports, so you re-enter what you use and drop what you do not. You give up Command Mode and the notetaker; you gain EU processing by a German GmbH, a public AVV, Linux and formatting that works in German. **Start your 14-day Sprecho Pro trial without a credit card and run the checklist above today**, or read the full [Sprecho vs Wispr Flow comparison](/compare/sprecho-vs-wispr-flow) first. ## FAQ ### Can I import my Wispr Flow dictionary into Sprecho? No. Sprecho has no Wispr Flow import, and Wispr's help center documents no dictionary export or CSV option, so you re-enter the words by hand under Dictionary → Add new. For about 50 entries that takes five to six minutes, and it is a good moment to drop words you no longer use. ### Do my Wispr Flow snippets transfer to Sprecho? Not automatically. You re-create each one in Sprecho under Snippets → Add new with a Trigger phrase and the text under Expands to. Sprecho snippets also support the variables {date}, {time} and {clipboard}, and can be shared with your team. ### Does Sprecho have a "Hey Flow" phrase or a Command Mode? No. Wispr's own docs say "Hey Flow" does not start Command Mode during ordinary dictation either; the shortcut does. Sprecho always starts from the dictation key, either held for push-to-talk or tapped once for hands-free lock mode, and it has no voice editing of existing text. It does support punctuation and layout commands in German, English and Portuguese, plus "scratch that" / "Streich das". ### Is there a free Sprecho plan after the trial? Yes. Every account starts with a 14-day Pro trial without a credit card and then continues on Free with 2,000 words per week, resetting every Monday at 00:00 UTC. Pro costs €12.99 per month or €10.99 per month billed yearly, both brutto including 19 % VAT. ### How long does switching from Wispr Flow to Sprecho take? About 20 minutes for one person with around 50 dictionary words and 10 snippets: three minutes to install, two for the hotkey, six for the dictionary, four for snippets, two each for formatting and retention, and one for a test. Larger dictionaries take longer, but you can add the rest over time while auto-learning fills gaps. ### Can my whole team switch to Sprecho together? Yes. Create a team under Team → Add your team, invite colleagues by email or CSV/Excel, and build one shared dictionary and shared snippets with Share with team switched on. Shared entries are included in Pro; Enterprise adds the admin dashboard, licence management, SSO/SAML and MDM, and every business customer gets an AVV under Art. 28 GDPR. ### What happens to my Wispr Flow subscription when I switch? You cancel it with the provider that bills you: Stripe for desktop or web purchases, Apple for App Store purchases, Google Play for Android. According to Wispr's cancellation article, paid access runs to the end of the billing period, there is no automatic prorated refund, and deleting your Wispr account does not cancel a store subscription; check their terms for refunds. ### Is Sprecho available on Mac, Windows and Linux? Yes. Sprecho runs natively on Windows 10 and 11, macOS on Apple Silicon and Intel, and Linux, plus iOS and Android apps, with one licence covering three devices. Wispr Flow offers Mac, Windows, iPhone and Android but no Linux app. --- ## https://sprecho.ai/blog/is-wispr-flow-gdpr-compliant — Is Wispr Flow GDPR-Compliant? What German Users and Companies Need to Know (2026) > Is Wispr Flow GDPR-compliant? US data processing, DPA only on Growth/Enterprise, CLOUD Act, privacy settings and § 203 StGB explained for German users. *Published: 2026-11-20T18:03:07.563000Z* *Tags: wispr-flow, gdpr, privacy, voice-dictation, germany, legal, comparison* **Is Wispr Flow GDPR-compliant?** For a private user dictating non-confidential English text, Wispr Flow can be used lawfully with care. For a German company, a practice or a law firm that dictates personal data, the honest answer is **not without significant extra work, and for many professions not at all**: Wispr AI, Inc. is a US company, it states that it "processes and stores customer data in the United States", and its data processing agreement is only listed on the Growth and Enterprise plans. *Last verified: October 2026.* **Wispr Flow** is an AI voice-dictation app by Wispr AI, Inc. (San Francisco) that turns speech into polished text in any app on Mac, Windows, iPhone and Android. **Sprecho** is a GDPR-native dictation app by Melo Designer GmbH from Lastrup, Germany, that processes speech exclusively on servers in the EU. This guide walks through Wispr Flow the way a data protection officer (DPO, in German *Datenschutzbeauftragter*) would: jurisdiction, contracts, sub-processors, settings, certifications and the professions for which the question is already decided. ## The short answer: is Wispr Flow GDPR-compliant? **Wispr Flow is not designed around the GDPR, and Wispr AI, Inc. does not claim GDPR compliance on its website.** It can still be part of a lawful setup, but only if you close several gaps yourself. - **US company, US processing:** the [security FAQ](https://docs.wisprflow.ai/articles/3467817258-security-and-compliance-faq) states verbatim that "Wispr Flow is a cloud service, not an on-premise deployment, and it processes and stores customer data in the United States." There is no EU data residency option. - **No DPA on Free and Pro:** on the [pricing page](https://wisprflow.ai/pricing) (checked 5 October 2026) the row "MSA & DPA" is ticked only for Growth and Enterprise. A business using Free or Pro for personal data has no Art. 28 GDPR contract. - **No transfer mechanism named publicly:** the [privacy policy](https://wisprflow.ai/privacy-policy) (last updated 19 August 2026) confirms transfers to the US but names neither the EU Standard Contractual Clauses nor the EU-US Data Privacy Framework. - **Good security, different law:** SOC 2 Type II, ISO/IEC 27001:2022 and HIPAA readiness are real strengths, but they are security and US health-law attestations, not proof of GDPR compliance. - **Verdict:** acceptable for private, non-confidential use; a heavy compliance project for ordinary businesses; **not suitable for professionals bound by § 203 StGB** or for most public bodies. ## Who runs Wispr Flow and where does your data go? **Wispr Flow is run by Wispr AI, Inc., a Delaware corporation based in San Francisco, and your audio is processed in the United States.** These are the facts a GDPR assessment starts from. - **Legal entity:** Wispr AI, Inc., a US corporation. We found no German or EU subsidiary on its website. - **Processing location:** the United States, per Wispr's own [security and compliance FAQ](https://docs.wisprflow.ai/articles/3467817258-security-and-compliance-faq). The [Trust Center](https://trust.wispr.ai) lists Amazon Web Services among its sub-processors; independent research in May 2026 placed primary processing in AWS us-east-1 (Virginia). - **Cloud-only:** there is no offline mode. Every dictation is sent to Wispr's servers to be transcribed and formatted. - **Encryption:** "Wispr encrypts service data in transit and at rest. It is not end-to-end encrypted in the strict sense, because the service must process audio to produce a transcript." (security FAQ) - **Growth and funding:** Wispr raised a **$280 million Series B** led by Menlo Ventures, announced on 17 August 2026 at a $2 billion valuation, and says it is moving "beyond dictation" into meetings and enterprise ([TechCrunch](https://techcrunch.com/2026/08/17/wispr-raises-280m-at-2b-valuation-as-it-looks-beyond-dictation/)). More enterprise features usually means more data per user, not less. ## Server location is not GDPR compliance: the CLOUD Act problem **The decisive question is not where the server stands but which legal system can force the provider to hand over data.** Under the US CLOUD Act of 2018, US authorities can compel a US company to disclose data in its "possession, custody, or control", wherever that data is stored. | Setup | US Jurisdiction? | GDPR-Safe? | |---|---|---| | US company + US servers | Yes | ❌ No | | US company + EU servers | **Yes (CLOUD Act)** | **❌ No** | | EU company + EU servers | No | **✅ Yes** | Wispr Flow sits in the first row: a US company processing in the US. That has three consequences for a German organisation. - **Third-country transfer:** every dictation is a transfer of personal data to the US under Art. 44 to 49 GDPR, so you need a valid transfer mechanism before the first word is spoken. - **Schrems II:** in its judgment of 16 July 2020 (C-311/18) the Court of Justice of the EU struck down the Privacy Shield and required exporters to check, case by case, whether US surveillance law undermines the Standard Contractual Clauses. That check is a **transfer impact assessment**. - **Data Privacy Framework:** the EU-US Data Privacy Framework would simplify transfers to certified US companies, but Wispr's [privacy policy](https://wisprflow.ai/privacy-policy) of 19 August 2026 does not mention it. Without it, you rely on whatever transfer terms sit in a DPA you may not have. - **Professional secrecy:** for lawyers, doctors and tax advisors the question is not only GDPR but criminal law (§ 203 StGB). See the decision table below. ![Server room corridor symbolising where dictation data is processed and stored](https://api.sprecho.ai/media/2026/10/1791204637-c56a49b0.jpg) ## Wispr Flow against a GDPR checklist **Measured against what a German DPO asks for, Wispr Flow passes on security controls and misses on contracts, transparency and data location.** The table compares what Wispr publishes with what Sprecho publishes, as of 5 October 2026. | Requirement | Wispr Flow (as published) | Sprecho | |---|---|---| | Company and jurisdiction | Wispr AI, Inc., Delaware (US) | Melo Designer GmbH, Lastrup (Germany) | | Processing location | ❌ United States | ✅ Germany and the Netherlands | | EU data residency option | ❌ None offered | ✅ Default, no option needed | | DPA / AVV (Art. 28 GDPR) | ⚠️ Growth and Enterprise only ("MSA & DPA") | ✅ Every business customer, every plan, public PDF | | Public sub-processor list | ⚠️ 5 names public, full list via Trust Center access | ✅ Full list public with role and region | | Transfer mechanism named | ❌ Not in the privacy policy | ✅ Not needed: no third-country transfer | | EU representative (Art. 27) | ❌ None found | ✅ Not needed: EU company | | Data protection officer named | ❌ None named in the privacy policy | ✅ Alex Melo, TÜV SÜD-certified | | Training on customer data | ⚠️ Opt-out setting; "never trained" promise on Enterprise only | ✅ Audio and transcripts never used for training; optional anonymised text patterns via the 'Help improve Sprecho' switch (on by default, can be turned off) | | Zero retention possible | ✅ Sharing off + Dictation Cloud Storage off | ✅ "Delete immediately after pasting", enforced server-side | | Third-party AI providers | ⚠️ Yes, under zero-retention agreements | ✅ None; own models on EU GPUs | | Security certifications | ✅ SOC 2 Type II, ISO/IEC 27001:2022 (company audits) | ⚠️ No own SOC 2 or ISO audit yet; hosting at ISO 27001-certified STRATO, TOMs in the AVV | | US health law (HIPAA) | ✅ HIPAA-ready with a signed BAA | ➖ Not relevant for German practices; Art. 9 GDPR and § 203 StGB apply instead | - **Where Wispr Flow is ahead:** formal third-party audits of the company itself. SOC 2 Type II and ISO 27001 are serious work; Sprecho relies on its ISO 27001-certified hoster and has not completed its own SOC 2 audit. - **Where Sprecho is ahead:** everything that follows from being an EU company: no transfer, no transfer impact assessment, an AVV on every plan and a public sub-processor list. - **What the table cannot show:** the content of Wispr's DPA, which is not public. Ask for it before you decide; read its transfer clause first. ## Does Wispr Flow offer an AVV or DPA, and who are the sub-processors? **Wispr offers a DPA, but on the public pricing page only on the Growth and Enterprise plans, and the full sub-processor list sits behind an access request.** For individuals on Free or Pro who use the app for work, that is the first gap. ### The DPA - **Plan restriction:** "MSA & DPA" is shown for Growth ($23 per user per month, $18 billed annually) and Enterprise (custom, annual only); the Free and Pro columns are empty. - **Wording:** the [security FAQ](https://docs.wisprflow.ai/articles/3467817258-security-and-compliance-faq) advises: "Review your DPA and its transfer terms if your organization has regional requirements." The DPA text itself is not published. - **What you need from it:** processor obligations under Art. 28(3) GDPR, the transfer mechanism (Standard Contractual Clauses), a sub-processor change procedure with a right to object, and deletion at the end of the contract. ### The sub-processor list - **Publicly visible:** the [Trust Center](https://trust.wispr.ai) names five sub-processors: Amazon Web Services, Baseten, Vercel, OpenAI and Supabase. No regions are shown. - **Behind access:** the security FAQ says "the Trust Center holds the current subprocessor list and supporting documents"; the detail requires an access request. - **Independent analyses count more:** a May 2026 review counted around fifteen vendors, including further AI providers and analytics tools. The authoritative version is the one Wispr gives you under your DPA. - **Third-party AI providers:** Wispr says it "always maintains zero data retention agreements with all third-party AI providers" ([data controls page](https://wisprflow.ai/data-controls)). Its [privacy policy](https://wisprflow.ai/privacy-policy) adds that content given to such providers "is generally deleted within 30 days, subject to the provider's applicable retention practices and legal obligations". Ask which statement your contract follows. ## Wispr Flow's privacy settings: what they do and what they do not **Wispr Flow has two separate controls for your dictations, plus Context Awareness and Notetaker settings, and you need all of them configured to reach zero retention.** Wispr's own wording, quoted from its [data controls page](https://wisprflow.ai/data-controls): ### Context Awareness - **What Wispr says:** "If 'Context Awareness' is enabled in Settings > Data and Privacy, relevant text data from the active app window may be used to improve transcription accuracy." - **Text, not screenshots:** Wispr's wording describes text from the active window, not images. Claims that the feature takes screenshots are not backed by Wispr's documentation. - **GDPR relevance:** the text in the active window can include a client email or a patient record, which then travels to the US with the dictation. Turn it off for confidential work. ### Model training sharing and Dictation Cloud Storage - **Training sharing (formerly "Privacy Mode"):** "This setting was previously called 'Privacy Mode.' If you had Privacy Mode enabled, your data is not shared for model training." Wispr may use shared data to improve its own models. - **Dictation Cloud Storage:** "Dictation Cloud Storage controls whether Wispr stores your transcription data (transcripts, audio, and dictation history) on Wispr's servers." - **Zero retention needs both:** Wispr's [HIPAA announcement](https://wisprflow.ai/post/hipaa-is-here) of 6 August 2025 states "Data sharing off, paired with dictation cloud storage off, guarantees zero data retention for dictation data." - **Defaults:** independent reviews report that storage is on by default for standard accounts. Check both switches on every device and, on Enterprise, enforce them for the whole organisation. - **What zero retention does not change:** the audio is still transmitted to and processed in the US. Retention settings reduce what is kept; they do not change where processing happens. ### Flow Notetaker - **Storage:** "Notetaker transcripts are stored on Wispr's private cloud for cross-device syncing and resumption of meetings." Retention is configurable in Settings > Data and Privacy. - **Audit scope:** the security FAQ states that "Meeting Notetaker launched after those audit periods and was not assessed in the SOC 2 Type I, the SOC 2 Type II, or the ISO 27001 certification audit." - **Other people's voices:** meeting recording processes the speech of every participant, which requires a legal basis and information for each of them under Art. 13 GDPR. ## SOC 2, ISO 27001 and HIPAA: why they do not equal GDPR **Certifications prove that a provider runs good security controls; they do not answer GDPR questions about jurisdiction, transfers and legal basis.** Wispr Flow lists SOC 2 Type II, ISO/IEC 27001:2022 and HIPAA readiness with a signed BAA. - **SOC 2:** an AICPA attestation (a US auditing standard) about security, availability and confidentiality controls. It says nothing about CLOUD Act exposure. - **ISO/IEC 27001:** an international information security management standard. Useful evidence for Art. 32 GDPR (security of processing), but only one article of many. - **HIPAA:** a US health privacy law. A BAA is the US counterpart of an AVV, but it does not satisfy Art. 28 GDPR, Art. 9 GDPR or § 203 StGB. - **The Delve episode:** Wispr's earlier SOC 2 and ISO reports were produced with the compliance vendor Delve, which faced public allegations of templated reports in March 2026. Wispr engaged new auditors for a fresh audit, and its security FAQ now lists a current SOC 2 Type II. - **Takeaway:** treat the certifications as a strong Art. 32 argument and nothing more. ## What a German DPO needs before approving Wispr Flow **Before Wispr Flow can be used with personal data in a German organisation, the DPO needs a contract, a transfer assessment and enforced settings.** A realistic checklist: 1. **Signed DPA (AVV) under Art. 28 GDPR:** in practice this means the Growth or Enterprise plan. 2. **Full sub-processor list** with role and location of each vendor, plus the change notification procedure. 3. **Transfer mechanism:** the Standard Contractual Clauses in the DPA, since no Data Privacy Framework participation is referenced in the privacy policy. 4. **Transfer impact assessment** following Schrems II, documenting why US access risks are acceptable for the data you dictate. 5. **TOMs (Art. 32 GDPR):** request the SOC 2 Type II report and the ISO certificate from the Trust Center. 6. **Data protection impact assessment (Art. 35 GDPR)** if health data, employee data or other high-risk data is dictated. 7. **Enforced settings:** training sharing off, Dictation Cloud Storage off, Context Awareness off, Notetaker restricted, ideally enforced org-wide. 8. **Records of processing (Art. 30 GDPR)** and employee information about what the tool sends where. For a team of five, that is several days of legal and IT work before the first dictation, and it has to be repeated when Wispr changes its sub-processors. ## Who can use Wispr Flow in Germany? **Whether Wispr Flow is acceptable depends on who dictates what.** The decision table summarises where the line runs. | User type | Wispr Flow | Note | |---|---|---| | Private individual, own notes and messages | ✅ Usable | Household exemption (Art. 2(2)(c) GDPR) for purely personal use; check the privacy settings | | Freelancer, non-confidential English content | ⚠️ With care | No DPA on Pro; avoid client personal data | | Company team dictating customer or employee data | ⚠️ Only with Growth/Enterprise | DPA, transfer impact assessment and enforced settings needed | | Lawyers (§ 203 StGB, § 43e BRAO) | ❌ Not advisable | US provider with US processing; secrecy protection abroad must be comparable | | Doctors and therapists (§ 203 StGB, Art. 9 GDPR) | ❌ Not advisable | Health data plus professional secrecy; a HIPAA BAA does not replace this | | Tax advisors (§ 203 StGB, [§ 62a StBerG](https://www.gesetze-im-internet.de/stberg/__62a.html)) | ❌ Not advisable | Same outsourcing rules as for lawyers | | Public sector and municipalities | ❌ Usually excluded | Citizen data, procurement and state data protection rules | ### Private use - **Lowest risk:** dictating your own shopping list or private messages falls largely outside the GDPR. - **Still worth configuring:** turn off training sharing and cloud storage if you do not want your dictations kept in the US. ### Lawyers, doctors and tax advisors - **Criminal law, not only data protection:** § 203 StGB makes disclosing client or patient secrets a criminal offence. Outsourcing is allowed only under strict conditions, including a contractual secrecy obligation for the service provider. - **Services abroad:** § 43e(4) BRAO allows lawyers to use service providers abroad only if the protection of secrets there is comparable to Germany. A US provider subject to the CLOUD Act is hard to justify. - **Health data:** dictated findings are special-category data under Art. 9 GDPR. Read our guides for [lawyers and law firms](https://sprecho.ai/blog/gdpr-speech-recognition-for-lawyers-law-firms) and the [Sprecho healthcare solution](https://sprecho.ai/solutions/healthcare). ![Law office desk with case files bound by professional secrecy under section 203 StGB](https://api.sprecho.ai/media/2026/10/1791204637-c83fcf37.jpg) ### Public sector - **Data sovereignty requirements:** authorities generally need EU processing, a German-law contract and transparent sub-processors. - **Procurement:** annual US contracts in USD with custom terms rarely fit public tenders. See [Sprecho for government and municipalities](https://sprecho.ai/solutions/government). ## Where Wispr Flow is genuinely strong **Wispr Flow is one of the most polished dictation products on the market, and a fair assessment says so.** - **Speed and polish:** very fast English dictation with filler removal and self-correction handling. - **Platforms:** native apps for Mac, Windows, iPhone and Android (Android since 23 February 2026). - **Features:** Command Mode for voice editing, hands-free dictation, snippets, a personal dictionary and Flow Notetaker. - **Security programme:** SOC 2 Type II, ISO/IEC 27001:2022, HIPAA with a BAA and a Trust Center. - **Free tier:** 2,000 words per week on desktop and 1,000 on mobile. ## The GDPR-native alternative: Sprecho **Sprecho closes exactly the gaps that make Wispr Flow hard to approve: EU company, EU processing, a public AVV and a public sub-processor list.** - **EU company:** Melo Designer GmbH, Alter Schützenweg 14, 49688 Lastrup, Germany. Data protection officer: Alex Melo, TÜV SÜD-certified. - **EU processing:** application, database and storage on STRATO GmbH servers in Germany; GPU speech recognition and formatting on Media Trooper in Germany and the Netherlands; off-site backups at Hetzner in Nuremberg. No US sub-processor in the dictation path and no third-party AI provider. - **AVV on every plan:** the data processing agreement under Art. 28 GDPR applies to every business customer, Free, Pro or Enterprise, and is available as a PDF. Details in [GDPR and your data processing agreement](https://sprecho.ai/help/privacy-and-security/gdpr-and-your-data-processing-agreement). - **Public sub-processor list:** every vendor with role and region at [where Sprecho runs and who processes your data](https://sprecho.ai/help/privacy-and-security/where-sprecho-runs-and-who-processes-your-data), with 30 days' notice before changes. - **Real zero retention:** Settings → Privacy → Data Retention → "Delete immediately after pasting" is enforced on the server before any row is written. - **Honest limits:** Sprecho is also cloud-based (no offline mode), and its ISO 27001 certification sits with the hoster STRATO, not with Sprecho itself; there is no SOC 2 report yet. - **Pro:** €10.92 net per month (€12.99 incl. 19 % VAT), or €9.24 net per month billed yearly (€110.82 net per year). - **Enterprise:** €21.84 net per seat per month, or €18.47 net billed yearly (€221.64 net per year), with admin dashboard, SSO/SAML, MDM and audit-log export; on-premise on request. - **For comparison:** Wispr Flow Pro is $15 per user per month or $12 billed annually; German visitors see the same digits in EUR, and Wispr states it adds no VAT (German businesses book reverse-charge VAT). - **Try it:** 14-day Pro trial without a credit card, plus a free plan with 2,000 words per week. See [pricing](https://sprecho.ai/pricing) and the full [Sprecho vs Wispr Flow comparison](https://sprecho.ai/compare/sprecho-vs-wispr-flow). ## Conclusion **Is Wispr Flow GDPR-compliant? Not out of the box.** Wispr AI, Inc. processes and stores data in the United States, lists its DPA only on Growth and Enterprise, keeps the full sub-processor list behind an access request and does not name a transfer mechanism in its privacy policy. Its certifications are real, but they answer security questions, not jurisdiction questions. Private users can configure it sensibly; businesses need a DPA, a transfer impact assessment and enforced settings; lawyers, doctors, tax advisors and public bodies should not route confidential dictation through it. - **If you need dictation that your DPO can approve in an afternoon:** **try Sprecho free for 14 days, no card required, with EU-only processing and an AVV on every plan**. - **Read next:** our [Wispr Flow alternatives for Germany](https://sprecho.ai/blog/wispr-flow-alternatives-germany), [where your voice is stored with AI dictation](https://sprecho.ai/blog/where-is-my-voice-stored-ai-dictation) and [is Otter.ai GDPR-compliant?](https://sprecho.ai/blog/is-otter-ai-gdpr-compliant). - **For law firms:** see [Sprecho for legal](https://sprecho.ai/solutions/legal). ## FAQ ### Is Wispr Flow GDPR-compliant? Not out of the box. Wispr AI, Inc. is a US company that processes and stores customer data in the United States, lists its DPA only on the Growth and Enterprise plans and does not name a transfer mechanism in its privacy policy. Private users can use it with sensible settings, but businesses need a DPA, a transfer impact assessment and enforced settings before dictating personal data. ### Where does Wispr Flow store my data? In the United States. Wispr's security FAQ states that Wispr Flow "processes and stores customer data in the United States", and Amazon Web Services is among the sub-processors listed in its Trust Center. There is no EU data residency option. ### Does Wispr Flow offer an AVV or DPA? Yes, but according to the pricing page checked on 5 October 2026 the "MSA & DPA" row is ticked only for the Growth and Enterprise plans. Free and Pro users who dictate personal data for work therefore have no Art. 28 GDPR contract, and the DPA text itself is not public. ### Are SOC 2, ISO 27001 and HIPAA enough for the GDPR? No. SOC 2 and ISO/IEC 27001 prove good security controls and support Art. 32 GDPR, and HIPAA is US health law. None of them addresses third-country transfers, the US CLOUD Act or the Schrems II ruling of 16 July 2020, which are the core GDPR issues for a US provider. ### Can lawyers, doctors or tax advisors use Wispr Flow under § 203 StGB? It is not advisable. § 203 StGB makes disclosing client or patient secrets a criminal offence, and § 43e(4) BRAO only allows service providers abroad if secrecy protection there is comparable to Germany. A US provider processing in the US and subject to the CLOUD Act is very hard to justify, and a HIPAA BAA does not change that. ### Does Wispr Flow's Context Awareness read my screen? According to Wispr's data controls page, when Context Awareness is enabled "relevant text data from the active app window may be used to improve transcription accuracy". Wispr describes text, not screenshots. For confidential work you should still turn it off, because that text can contain client or patient data. ### Does Wispr Flow train on my dictations? Only if you leave the data-sharing setting (formerly Privacy Mode) on; Wispr says its third-party AI providers cannot train on your data whatever the setting. For zero retention Wispr says you need data sharing off and Dictation Cloud Storage off, and only the Enterprise plan carries the promise that models are never trained on your data. ### What does Sprecho do differently? Sprecho is run by Melo Designer GmbH in Germany and processes dictations only on servers in Germany and the Netherlands, without any third-party AI provider. Every business customer gets an AVV under Art. 28 GDPR on every plan, the full sub-processor list is public, and "Delete immediately after pasting" is enforced on the server. --- ## https://sprecho.ai/blog/wispr-flow-280-million-series-b-what-it-means-for-european-users — Wispr Flow Raised $280 Million: What the Series B Means for European Users (2026) > Wispr Flow raised $280M at a $2B valuation in August 2026. What the money funds, what changed in 2026, and the GDPR questions European buyers should still ask. *Published: 2026-11-20T18:03:06.338000Z* *Tags: wispr-flow, gdpr, voice-dictation, germany, comparison, privacy, pricing* **Wispr Flow's Series B** closed on 17 August 2026: Wispr AI, Inc. raised **$280 million** led by Menlo Ventures at a **$2 billion valuation**, bringing its total funding to $361 million. This article sets out what was announced, what Wispr says it will build with the money, what usually follows a round of this size, and which questions stay exactly the same for European and German buyers after the cheque has cleared. *Last verified: October 2026.* **Wispr Flow** is a cloud voice-dictation app for Mac, Windows, iPhone and Android, made by Wispr AI, Inc., a Delaware corporation based in San Francisco. **Sprecho** is a voice-dictation app for Windows, macOS, Linux, iOS and Android, made by Melo Designer GmbH in Lastrup, Germany, and processed on servers in Germany. We make Sprecho, so read this as a competitor's analysis; every Wispr Flow fact below cites Wispr's own pages or TechCrunch. ## Wispr Flow's Series B at a glance **Wispr raised $280 million in a Series B led by Menlo Ventures, announced on 17 August 2026, at a $2 billion valuation and a total of $361 million raised to date**, according to [TechCrunch's report](https://techcrunch.com/2026/08/17/wispr-raises-280m-at-2b-valuation-as-it-looks-beyond-dictation). The previous round closed less than ten months earlier. - **Amount:** $280 million, Series B. - **Lead investor:** Menlo Ventures. - **Valuation:** $2 billion, roughly three times the about $700 million reported for the November 2025 Series A extension. - **Total raised:** $361 million since founding, so the Series B alone is more than three quarters of all money Wispr has taken in. - **Stated direction:** "beyond dictation", into meeting notes, a proprietary speech model and larger team accounts. | Date | Event | Detail | |---|---|---| | November 2025 | Series A extension | About $25 million led by Notable Capital, reported at roughly $700 million post-money | | 23 February 2026 | Android app | Fourth platform after Mac, Windows and iPhone | | 27 March 2026 | New audit partners | Drata and A-LIGN engaged after the Delve episode | | 5 August 2026 | Flow Notetaker (Mac) | "Our first product beyond dictation", meeting recorder without a bot | | **17 August 2026** | **Series B** | **$280 million, Menlo Ventures, $2 billion valuation, $361 million total** | | 28 August 2026 | Growth plan | Self-serve team plan bought from the admin portal | | 15 September 2026 | Notetaker on Windows | 21 transcription languages, Otter.ai transcript import | | 17 September 2026 | Canto | "The first speech model we've built ourselves", rolled out to all users | ## Who invested, and what a $2 billion valuation signals **The round combines returning backers with a group of new growth investors, and the valuation tells you investors expect Wispr to become much more than a dictation utility.** A company priced at $2 billion has to grow into a far larger revenue base than a single-purpose app usually reaches. - **Lead:** Menlo Ventures. - **Returning investors:** Notable Capital, NEA, Neo, 8VC and MVP. - **New investors named by TechCrunch:** Acrew, Forerunner, Goodwater, Peak XV, Together Fund and PLUS Capital. - **Competitors named in the coverage:** Willow, Monologue, Aqua and Superwhisper for dictation; Granola, Fireflies and Read AI for meeting notes. What the investor list does **not** contain is just as relevant for a European reader: - **No European entity announced:** the coverage mentions go-to-market teams in India and the U.K., not an EU subsidiary. - **No EU data-residency plan announced:** nothing in the funding news or the release notes points to EU hosting. - **No revenue or user numbers disclosed:** the article gives no figures, so any "X million users" claim you read elsewhere is not from this announcement. ## What Wispr says it will build with the money **Wispr's own messaging is "beyond dictation": a meeting notetaker, its own speech model, and team plans that sell to whole organisations.** All three had launched or been announced within six weeks of the round, so they are concrete, not roadmap slides. ![Abstract rising growth chart on an office monitor, symbolising Wispr Flow's 2026 funding round](https://api.sprecho.ai/media/2026/10/1791204637-25459c88.jpg) ### Flow Notetaker: meetings without a bot **Flow Notetaker launched on Mac on 5 August 2026 and on Windows on 15 September 2026, per [Wispr's release notes](https://wisprflow.ai/whats-new).** It records Zoom, Google Meet, Teams and Slack huddles from the desktop, without a bot joining the call. - **Languages:** 21, including German, Spanish, French, Chinese, Japanese, Hindi and Arabic. - **Output:** transcripts with speaker names and summaries. - **Migration hook:** transcripts can be imported from Otter.ai with speaker labels. - **Storage:** "Notetaker transcripts are stored on Wispr's private cloud for cross-device syncing and resumption of meetings" ([data controls](https://wisprflow.ai/data-controls)). ### Canto: Wispr's first proprietary speech model **Canto was announced on 17 September 2026 and now runs for every Flow user "with nothing to switch on".** Wispr reports the lowest word error rate of every model it tested against Google, OpenAI, AssemblyAI and Deepgram. - **Evaluation set:** about 10 hours of **English-language** Flow dictations from more than 2,300 speakers, per [the Canto research post](https://wisprflow.ai/canto). - **German:** no German results are published; Wispr says it aims to reach "the same standard across more languages" with a successor model. - **Error-rate promise:** TechCrunch quotes the company as saying Canto will reduce error rates "from 30% to less than 10%". ### Growth plan and the push into organisations **On 28 August 2026, eleven days after the round, Wispr introduced the self-serve Growth plan.** Teams buy it from the admin portal, choosing dictation only or a bundle with unlimited Notetaker. - **Growth, dictation only:** $23 per user per month, or $18 billed annually ([pricing page](https://wisprflow.ai/pricing), 5 October 2026). - **Growth with unlimited Notetaker:** $33 per user per month, or $26 billed annually. - **Growth adds:** SSO with SAML, "Org-wide HIPAA enforcement with a signed BAA" and admin control over model training for the whole team. - **Enterprise:** custom pricing, annual only, with SCIM, audit logs, MDM, domain capture and "Model training locked off". ## What changed in Wispr Flow in 2026 **2026 turned Wispr Flow from a dictation app into a small product suite**: a fourth platform, a re-run security audit, a team plan, a meeting product and an in-house model. The table summarises each change and what it means for a buyer in the EU. | Change | Date | What it means for European users | |---|---|---| | Android app | 23 February 2026 | ✅ Same app on all four major platforms; Linux is still not supported | | New SOC 2 audit with A-LIGN | Partners engaged 27 March 2026 | ✅ Clean SOC 2 Type I in April 2026; SOC 2 Type II now listed; a US attestation, not a GDPR assessment | | Flow Notetaker | 5 August 2026 (Mac), 15 September 2026 (Windows) | ❌ Not covered by the SOC 2 or ISO 27001 audits; transcripts stored in Wispr's cloud | | Growth plan | 28 August 2026 | ✅ Self-serve SSO and admin controls; prices in USD, no VAT added (reverse charge for German businesses) | | Canto speech model | 17 September 2026 | ❌ Benchmarked on English only; no German figures published | | Data processing location | Unchanged | ❌ "Processes and stores customer data in the United States" | ## Where Wispr Flow is genuinely strong **A well-funded competitor ships faster, and Wispr Flow is proof.** Pretending otherwise would not help anyone choose; these are real advantages. - **Release pace:** five significant launches between February and September 2026. - **English dictation:** very fast and polished, with an in-house model trained on real Flow usage. - **Platform coverage:** native apps for Mac, Windows, iPhone and Android. - **Voice editing:** Command Mode for instructions on existing text, with a hands-free lock. - **Meetings:** a bot-free Notetaker that Sprecho does not offer. - **US-style compliance:** SOC 2 Type II, ISO/IEC 27001:2022 and HIPAA with a signed BAA on Growth and Enterprise. - **Free tier:** 2,000 words per week on desktop and 1,000 on mobile, according to the pricing page. ## What usually follows a round like this **Nothing about prices has been announced, but large growth rounds tend to bring more tiers, a sharper enterprise focus and more products built on usage data.** Some of that is already visible; the rest is pattern, not prediction. - **More tiers (already happening):** Growth replaced the earlier Teams plan, and the Notetaker bundle costs $10 more per user per month than dictation-only Growth. - **Introductory offers that end:** the pricing page currently offers teams on dictation-only Growth free Notetaker access "through October 31"; plan for the bundle price after that. - **Enterprise focus:** SSO, SCIM, MDM and audit logs move up the feature list; expect more sales-led contracts. - **Geographic expansion:** India and the U.K. are named; the EU is not. - **More data-driven products:** a speech model trained on real Flow recordings, and admin settings that govern "model training for your whole team", show that usage data is a strategic asset. - **Price changes:** **not announced**. If you sign a yearly plan, the price for that term is what your order confirms. ## What the funding does not change for European users **$280 million buys engineers and sales teams; it does not move a Delaware company under EU law or its servers into the EU.** For a German buyer, six questions decide whether a dictation tool is usable, and none of the answers changed on 17 August 2026. - **US company + US servers:** US jurisdiction applies; not GDPR-safe without additional safeguards. - **US company + EU servers:** US jurisdiction still applies through the CLOUD Act; server location alone does not fix it. - **EU company + EU servers:** no US jurisdiction over the provider; the setup GDPR was written for. ### 1. Jurisdiction - **Entity:** Wispr AI, Inc. is a US corporation, so the CLOUD Act of 2018 lets US authorities demand data it controls, wherever that data sits. - **Transfers:** since Schrems II (CJEU C-311/18, 16 July 2020), EU-US transfers need a documented legal basis and a transfer impact assessment on the customer's side. - **EU representative:** we found no Art. 27 GDPR representative named by Wispr. ### 2. Data location - **Wispr's own words:** "Wispr Flow is a cloud service, not an on-premise deployment, and it processes and stores customer data in the United States" ([security and compliance FAQ](https://docs.wisprflow.ai/articles/3467817258-security-and-compliance-faq)). - **No EU option:** neither the FAQ nor the pricing page offers EU data residency on any plan. ### 3. Retention and controls - **Two separate switches:** the former "Privacy Mode" now controls sharing for model training; "Dictation Cloud Storage controls whether Wispr stores your transcription data (transcripts, audio, and dictation history) on Wispr's servers." - **Context Awareness:** when enabled, "relevant text data from the active app window may be used to improve transcription accuracy." - **Meeting audio:** the security FAQ says meeting audio is kept for 7 days after upload, then deleted. - **Third-party AI:** Wispr states it "maintains zero data retention agreements with all third-party AI providers". ### 4. Sub-processors - **Where the list lives:** "the Trust Center holds the current subprocessor list and supporting documents." - **Gap to check:** in May 2026 the public summary named five providers, while independent analyses counted around fifteen, including several US AI providers. Ask for the complete list before you sign. ### 5. Notetaker sits outside the audits ![Colleagues in a bright European meeting room, the setting where meeting recorders such as Flow Notetaker are used](https://api.sprecho.ai/media/2026/10/1791204637-b29cfbc2.jpg) - **Verbatim:** "Meeting Notetaker launched after those audit periods and was not assessed in the SOC 2 Type I, the SOC 2 Type II, or the ISO 27001 certification audit." - **Why it matters:** meetings carry other people's voices, so you need a legal basis for every participant, not just the person recording. - **Audit history:** Wispr's earlier reports were issued via Delve, the compliance vendor exposed in March 2026; Wispr engaged Drata and A-LIGN on 27 March 2026 ([Wispr's post](https://wisprflow.ai/post/new-independent-audit)), and a clean A-LIGN SOC 2 Type I followed in April 2026. ### 6. German specialist vocabulary - **Published evidence:** Canto's benchmark is English only, so there is no public figure for German legal, medical or tax terms. - **Practical test:** dictate your own Fachsprache (statutes, drug names, DATEV terms) in any tool you evaluate, ours included. ## Questions to ask any dictation vendor in 2026 **Use this checklist with Wispr Flow, with Sprecho and with anyone else**; the answers below are what each company publishes today. ### ![Wispr Flow logo](https://sprecho.ai/images/competitors/wispr-flow.svg) Wispr Flow vs ![Sprecho logo](https://sprecho.ai/sprecho-icon.svg) Sprecho: the answers side by side | Question | Wispr Flow (as published) | Sprecho | |---|---|---| | Which law governs the provider? | ❌ US (Delaware corporation) | ✅ German (Melo Designer GmbH) | | Where is voice processed and stored? | ❌ United States | ✅ Germany (STRATO), GPU inference Media Trooper (DE/NL) | | Is the sub-processor list public? | Behind the Trust Center | ✅ Public list in the help center | | Is there a DPA / AVV under Art. 28 GDPR? | DPA with transfer terms | ✅ Public AVV, German and English | | Can storage be switched off? | ✅ Dictation Cloud Storage setting | ✅ "Delete immediately after pasting", enforced server-side | | Is every product inside the audit scope? | ❌ Notetaker not assessed | No meeting recorder to scope | | German benchmark published? | ❌ No (Canto: English only) | ❌ No; test with the 14-day trial | | Price for one user, monthly | $15 Pro; € digits shown in Germany, no VAT added | €12.99 brutto (€10.92 netto) | | Invoice | US company, reverse charge for B2B | German GmbH invoice with VAT line | | Meeting notetaker | ✅ Notetaker (Mac, Windows) | ❌ No live recorder; audio file transcription | | Linux | ❌ | ✅ .deb and Fedora | - **Read the documents, not the headline:** the DPA, the sub-processor list and the audit scope matter more than the logo wall. - **Ask about defaults:** which privacy settings are on when a new user installs the app, before any admin intervenes. - **Ask about new products:** features launched after the last audit window are not covered by that report. ## Who should care most **Professionals bound by confidentiality law and public bodies are the readers for whom the funding changes nothing and the jurisdiction changes everything.** - **Lawyers:** § 203 StGB and § 43e BRAO set strict conditions for involving service providers; see [Sprecho for law firms](/solutions/legal). - **Doctors and clinics:** patient data is special-category data under GDPR Art. 9; see [Sprecho for healthcare](/solutions/healthcare). - **Tax advisors:** client data falls under § 203 StGB as well; see [Sprecho for tax consultants](/solutions/tax-consultant). - **Public sector:** authorities usually require EU processing and a German contract partner. - **English-first freelancers:** if your work is in English and not confidential, Wispr Flow's speed and Notetaker may simply suit you better. ## Sprecho's answer: EU processing and a German invoice **Sprecho does not try to match Wispr Flow feature for feature; it answers the six questions above differently.** The short version: a German company, German servers and documents a data protection officer can file. - **Company:** Melo Designer GmbH, Alter Schützenweg 14, 49688 Lastrup; data protection officer Alex Melo, TÜV SÜD-certified. - **Infrastructure:** STRATO servers in Germany, GPU inference by Media Trooper (DE/NL), off-site backups at Hetzner in Nürnberg; see [where Sprecho runs](/help/privacy-and-security/where-sprecho-runs-and-who-processes-your-data). - **Contract:** a public AVV under Art. 28 GDPR for business customers; see [GDPR and your DPA](/help/privacy-and-security/gdpr-and-your-data-processing-agreement). - **Retention:** "Delete immediately after pasting" is enforced on the server; see [turn off storage](/help/privacy-and-security/turn-off-transcript-and-audio-storage). - **Honest limits:** Sprecho is cloud-based with no offline mode, and it has no live meeting recorder. Prices are in euros, with VAT shown (see [pricing](/pricing)): - **Free:** 2,000 words per week, no card. - **Pro:** €12.99 per month or €10.99 per month billed yearly (€131.88 per year), brutto incl. 19 % VAT; netto €10.92 or €9.24. - **Enterprise:** €25.99 or €21.99 per seat per month brutto (€21.84 or €18.47 netto), with SSO/SAML, MDM, audit-log export, invoice billing and the AVV. - **Trial:** 14 days of Pro without a card. For the full feature-by-feature view, read [Sprecho vs Wispr Flow](/compare/sprecho-vs-wispr-flow) and our overview of [Wispr Flow alternatives for Germany](/blog/wispr-flow-alternatives-germany). If your question is mainly where voice data ends up, [where is my voice stored?](/blog/where-is-my-voice-stored-ai-dictation) walks through it. ## Conclusion **The $280 million Series B makes Wispr Flow a faster-moving, better-resourced product, but it does not change where your voice is processed or which law applies to it.** If you dictate in English and handle no confidential data, Wispr Flow's speed, Notetaker and Canto are strong reasons to try it. If you work in Germany under § 203 StGB, GDPR Art. 9 or a public-sector mandate, the deciding questions are jurisdiction, data location, retention, sub-processors and audit scope, and those answers are the same after the round as before it. **Test Sprecho for 14 days on your own German vocabulary, with processing in Germany and a German invoice: [compare Sprecho and Wispr Flow](/compare/sprecho-vs-wispr-flow).** ## FAQ ### How much did Wispr raise in its Series B? Wispr AI, Inc., the company behind Wispr Flow, raised $280 million in a Series B led by Menlo Ventures, announced on 17 August 2026. That brings its total funding to $361 million, according to TechCrunch. The previous round, a Series A extension, had closed less than ten months earlier. ### What is Wispr Flow's valuation? The Series B valued Wispr at $2 billion. That is roughly three times the about $700 million reported for its November 2025 Series A extension. Wispr has not published revenue or user numbers alongside the round. ### Who invested in Wispr Flow? Menlo Ventures led the Series B. Returning investors were Notable Capital, NEA, Neo, 8VC and MVP, and TechCrunch names Acrew, Forerunner, Goodwater, Peak XV, Together Fund and PLUS Capital as new backers. None of them changes the company's US domicile. ### What is Canto? Canto is Wispr's first in-house speech model, announced on 17 September 2026 and rolled out to all Flow users. Wispr reports the lowest word error rate against Google, OpenAI, AssemblyAI and Deepgram on about 10 hours of English Flow dictations from more than 2,300 speakers. No German benchmark has been published. ### Will Wispr Flow prices rise after the funding? No price increase has been announced. What has changed is the plan structure: the self-serve Growth plan launched on 28 August 2026 at $23 per user per month for dictation only and $33 with unlimited Notetaker, and the free Notetaker access for dictation-only Growth teams is advertised only through 31 October. ### Is Wispr Flow available in Germany? Yes. Anyone in Germany can download Wispr Flow for Mac, Windows, iPhone and Android, and German visitors see prices in euros. The service is operated by a US company that, in its own words, processes and stores customer data in the United States, and it offers no EU data residency. ### Does the funding change Wispr Flow's GDPR exposure? No. Jurisdiction and data location follow from the company and its infrastructure, not its funding: as a US corporation, Wispr is subject to the CLOUD Act, and EU-US transfers still need a legal basis and a transfer impact assessment after Schrems II (CJEU C-311/18). Professionals bound by § 203 StGB or handling health data under GDPR Art. 9 should evaluate it exactly as before the round. ### What does Sprecho offer instead? Sprecho is a dictation app by Melo Designer GmbH, processed on STRATO servers in Germany, with a public AVV under Art. 28 GDPR and a public sub-processor list. Pro costs €12.99 per month brutto or €10.99 per month billed yearly, with a 14-day trial without a card. Sprecho has no live meeting recorder and no offline mode, so it is an answer to the data-protection questions, not a copy of every Wispr feature. --- ## https://sprecho.ai/blog/wispr-flow-privacy-mode-cloud-storage-zero-retention-explained — Wispr Flow's Privacy Settings Explained: Why "Privacy Mode" Is Not Zero Data Retention > Wispr Flow privacy mode only stops training. Learn what Dictation Cloud Storage, Context Awareness and Notetaker keep, and how to set minimum retention. *Published: 2026-11-20T18:03:04.657000Z* *Tags: wispr-flow, privacy, gdpr, voice-dictation, germany, comparison* The **Wispr Flow privacy mode** is the setting most people switch on when they want their dictations to stay private, and it is also the setting most people misunderstand. Since August 2026 it no longer exists under that name: what used to be "Privacy Mode" is now the model-training switch, and storage of your audio and transcripts is governed by a separate control called Dictation Cloud Storage. This guide explains every Wispr Flow privacy setting in Wispr's own words, shows what each one does and does not stop, and walks you through configuring Wispr Flow for minimum data retention. *Last verified: October 2026.* - **Wispr Flow** is a cloud-based AI voice dictation app made by Wispr AI, Inc., a Delaware corporation based in San Francisco, available on Mac, Windows, iPhone and Android. - **Sprecho** is an AI voice dictation app made by Melo Designer GmbH in Lastrup, Germany, that processes dictation on GPU servers in Germany and the Netherlands (Media Trooper) and stores data on STRATO servers in Germany. - **Sources used:** Wispr's Data Controls page ([wisprflow.ai/data-controls](https://wisprflow.ai/data-controls), last updated 18 August 2026), the help-center articles "Security and privacy overview" ([docs.wisprflow.ai/articles/3467817258-security-and-compliance-faq](https://docs.wisprflow.ai/articles/3467817258-security-and-compliance-faq)) and "Manage data sharing, cloud storage and local history" ([docs.wisprflow.ai/articles/9609615338](https://docs.wisprflow.ai/articles/9609615338-private-cloud-sync-and-data-sharing-preferences-in-wispr-flow)), all read on 5 October 2026. ## The short answer: Privacy Mode stops training, not storage Turning off training (the former Privacy Mode) stops Wispr from using your content to improve its models, but it does not stop Wispr from storing your dictations. For storage you need a second switch, and some data categories are not covered by either switch. - **Training switch ("Improve the model for everyone"):** decides whether Wispr may use your audio, transcripts and edits to train or evaluate its own AI models. - **Dictation Cloud Storage:** decides whether Wispr keeps your transcripts, audio and dictation history on its servers. - **Independent controls:** Wispr states that "Model improvement and Dictation Cloud Storage are separate settings; changing one does not change the other." - **Outside both switches:** Notetaker meeting transcripts, Scratchpad notes on desktop, dictionaries, snippets, account data and usage statistics. - **Always in the cloud:** "Transcription always occurs on the cloud," and Wispr "processes and stores customer data in the United States." ## What "Privacy Mode" was, and what it is called now Privacy Mode was renamed: it is now the setting "Improve the model for everyone", and it only governs whether your data trains Wispr's models. Wispr's Data Controls page says so directly. - **The rename, verbatim:** "This setting was previously called 'Privacy Mode.' If you had Privacy Mode enabled, your data is not shared for model training. This behavior remains unchanged." - **What the setting allows when on:** "your data (i.e. audio, transcript, edits) may be used to evaluate, train, or improve AI models, by Wispr." - **Where to find it:** "Settings > Data and Privacy" in the Wispr Flow app. - **Default for individuals:** the help center states it "is on by default for Free and Pro, with the choice shown during onboarding, and you can turn it off." - **Default for organisations:** "On Growth and Enterprise it is off and administrators cannot re-enable it." - **Wispr's own warning:** "For confidential or privileged content, turn Improve the model for everyone off before recording. Turning it off does not stop cloud processing or change your storage settings." Older help-center articles, which now redirect to the current page, described **zero data retention** as the combination of Privacy Mode on and (Private) Cloud Sync off. The current documentation no longer uses the term for a user setting, but the logic still holds: privacy against training and privacy against storage are two different switches. ## The Wispr Flow privacy settings, one by one Wispr Flow has four controls that decide what happens to your dictation and meeting data, plus an export and deletion function. The table summarises them; the sections below quote Wispr's wording for each. | Setting | What it controls | Default (as documented or reported) | Effect when you change it | |---|---|---|---| | **Improve the model for everyone** (formerly Privacy Mode) | Use of dictation and Notetaker content (audio, transcripts, edits) to train or evaluate Wispr's models | On for Free and Pro; off and locked on Growth and Enterprise (Wispr help center) | Off: no training use; storage and cloud processing unchanged | | **Dictation Cloud Storage** | Storage of transcripts, audio and dictation history on Wispr's servers (dictation only) | Reported as on by default except under a HIPAA BAA (Voibe, April 2026); signing out resets it to on (Wispr help center) | Off: dictation data no longer stored in Wispr's cloud; turning off requires confirmation | | **Context Awareness** | Use of "relevant text data from the active app window" to improve accuracy | Not stated on the Data Controls page; reported as disabled by default (Voibe, April 2026) | Off: no active-window text beyond the always-on app name and textbox context | | **Notetaker transcript retention** | How long meeting transcripts stay in Wispr's cloud | Stored; meeting audio kept 7 days, then deleted (Wispr help center) | Retention period configurable; no option to opt out of storage | ![Padlock next to a closed laptop on an office desk, symbolising dictation data retention controls](https://api.sprecho.ai/media/2026/10/1791204608-c26fe560.jpg) ### Improve the model for everyone (training) This switch only governs training use, and Wispr documents its scope precisely. - **Covers meetings too:** it "does apply to Notetaker as well as dictation." - **Saves instantly:** the help center notes it "saves immediately, without confirmation." - **Fail-safe behaviour:** "If Flow cannot confirm a sharing preference, agreement or organization policy, it does not share dictation audio." - **Organisation effect:** because sharing stays off on Growth and Enterprise, "Wispr keeps no dictation transcripts for these organizations, so support cannot review transcript contents." ### Dictation Cloud Storage (retention) This is the switch that actually decides whether your dictations live on Wispr's servers. - **Wispr's definition:** "Dictation Cloud Storage controls whether Wispr stores your transcription data (transcripts, audio, and dictation history) on Wispr's servers." - **Why it exists:** it "enables current and future features that require persistent, cross-device access to your data." - **Scope limit:** it covers "dictation only, not Notetaker or Scratchpad." - **Sign-out trap:** "Signing out resets Dictation Cloud Storage to on," so check it every time you sign back in. - **BAA lock:** "A signed BAA locks both controls off." ### Context Awareness (active window text) Context Awareness reads text, not screenshots, according to Wispr's own wording. - **Verbatim:** "If 'Context Awareness' is enabled in Settings > Data and Privacy, relevant text data from the active app window may be used to improve transcription accuracy. You can turn this on/off at any time." - **Always on regardless:** "Wispr always uses information about the app you are dictating in (e.g. app name) to format messages" and "takes into account the surrounding textbox content to help determine correct capitalization and punctuation, and to edit text with Command Mode." - **Screenshots:** Wispr's pages describe text data, not images; independent reviews that speak of screenshots conflict with Wispr's own wording, so we do not repeat that claim. - **Related switch:** "Auto-add to Dictionary" lets Flow monitor "the text box where it pastes text to detect any edits you made"; it can be turned off in Settings > Personalization. ### Notetaker storage and retention Meeting data follows different rules from dictation data, and storage cannot be switched off. - **Data Controls page:** "Notetaker transcripts are stored on Wispr's private cloud for cross-device syncing and resumption of meetings. You can configure notetaker transcript retention in Settings > Data and Privacy." - **Help center:** "Dictation Cloud Storage does not govern them, and there is no customer setting to opt out of that storage." - **Meeting audio:** "kept for 7 days from upload to support troubleshooting and speaker attribution, then deleted; transcripts and summaries remain." - **Audit scope:** "Meeting Notetaker launched after those audit periods and was not assessed in the SOC 2 Type I, the SOC 2 Type II, or the ISO 27001 certification audit." ## Why Privacy Mode alone is not zero data retention Privacy Mode alone is not zero retention because it never controlled storage: your audio, transcripts and history stay on Wispr's servers unless Dictation Cloud Storage is also off, and several categories remain stored even then. Wispr lists them under "What privacy settings do not remove". - **Account data:** "Account information, preferences, and subscription details are kept to provide the service." - **Dictionaries and snippets:** "Content you create is stored and synced separately from dictated content." - **Diagnostics and usage statistics:** "Operational data such as word counts is distinct from dictation history"; the Data Controls page adds that Wispr "may collect usage statistics such as the number of words you have dictated, regardless of your data controls." - **Reported transcripts:** using Report or another feedback action "intentionally sends that content for investigation." - **Scratchpad notes:** "On Mac and Windows, notes sync to the cloud whatever your Dictation Cloud Storage setting." - **Notetaker recordings:** transcripts and summaries stay in Wispr's cloud with no opt-out. The feedback point deserves attention. According to the Data Controls page, rating a dictation or clicking thumbs up or down collects "data from the related session", which "may include audio recordings, transcripts, your edits, the app where you dictated", used "to diagnose issues, improve Wispr Flow, and develop and train our AI models." Wispr adds that sending feedback is optional. | Data type | Stored? | Where | Who can see it | |---|---|---|---| | **Dictation audio** | Only if Dictation Cloud Storage is on; always processed in the cloud | Wispr servers in the United States | Wispr; third-party AI providers process it under zero-retention agreements | | **Transcripts and dictation history** | Only if Dictation Cloud Storage is on | Wispr servers in the United States | You; Wispr staff only where access is permitted and logged | | **Feedback sessions (thumbs up/down, Report)** | Yes, when you send feedback | Wispr | "Authorized Wispr Flow staff and service providers bound by confidentiality obligations" | | **Notetaker transcripts and summaries** | Yes, no opt-out; retention configurable | "Wispr's private cloud" | You and your synced devices; Wispr as processor | | **Notetaker audio** | 7 days, then deleted | Wispr cloud | Wispr for troubleshooting and speaker attribution | | **Scratchpad notes (desktop)** | Yes, synced regardless of Dictation Cloud Storage | Wispr cloud | You; an organisation can turn Scratchpad off | | **Dictionary and snippets** | Yes | Stored and synced separately | You and your devices | | **Usage statistics (word counts)** | Yes, regardless of data controls | Wispr | Wispr (aggregate, no dictation contents per Wispr) | ## What happens at Wispr's third-party AI providers Wispr says its AI providers are bound by zero-retention and no-training agreements for dictation, with a documented exception for some Notetaker features. This applies whatever you choose in the training switch. - **No training by providers:** "Wispr maintains agreements that no third party AI providers can use your data for model training." - **Zero retention at providers:** "Wispr always maintains zero data retention agreements with all third-party AI providers. All third party AI models used in Flow Dictation are subject to these agreements." - **The exception:** "Certain features, such as briefs for Notetaker, rely on features from third party AI providers that are not supported under zero data retention." - **Who the providers are:** "the Trust Center holds the current subprocessor list and supporting documents"; restricted documents "may require approval and an NDA". Independent analyses have counted more providers than the public summary showed in May 2026. ## What a user cannot verify from the settings screen Even with every switch set to the strictest value, some things depend on contracts and audits you cannot inspect from the app. For an IT department or a data protection officer these are the open questions. - **Audio still leaves your device:** every dictation travels to Wispr's cloud for transcription; the settings control what is kept afterwards, not where processing happens. - **US processing:** "Wispr Flow is a cloud service, not an on-premise deployment, and it processes and stores customer data in the United States." Wispr AI, Inc. is a US company, so the US CLOUD Act (2018) applies to data in its control. - **Encryption keys:** "Wispr manages service encryption keys; customer-managed keys are not supported," and the service "is not end-to-end encrypted in the strict sense." - **Certification scope:** SOC 2 Type II and ISO/IEC 27001:2022 cover "Wispr Flow dictation"; Notetaker was not assessed. Wispr's earlier reports were issued via the compliance vendor Delve, which was exposed in March 2026; a new SOC 2 Type I by A-LIGN followed in April 2026, and the Trust Center now lists a current Type II. - **Transfer terms:** Wispr advises: "Review your DPA and its transfer terms if your organization has regional requirements." DPA requests run through the Trust Center. ## How to configure Wispr Flow for minimum retention, step by step To get as close to zero retention as Wispr Flow allows, you need to change several settings, not one. Follow these steps on Mac or Windows. 1. **Open the settings:** open Flow from the menu bar or system tray and go to Settings → Data & Privacy. 2. **Turn off training:** switch "Improve the model for everyone" off before you dictate anything confidential. 3. **Turn off Dictation Cloud Storage:** switch it off and accept the confirmation; this governs future storage, so handle data that is already stored in step 9. 4. **Turn off Context Awareness:** if you do not want active-window text used for accuracy. 5. **Turn off Auto-add to Dictionary:** under Settings → Personalization, if you do not want Flow watching your edits. 6. **Set Notetaker retention short or avoid Notetaker:** transcripts cannot be opted out of storage, only shortened. 7. **Do not send feedback on sensitive dictations:** thumbs up/down and Report send the session content. 8. **Re-check after every sign-in:** signing out resets Dictation Cloud Storage to on. 9. **Export, then delete:** request "Download your data" first if you need a copy, then delete stored records or request account deletion in Settings → Data and Privacy; Wispr warns that "Deleted records cannot be restored." - **For teams:** Growth ($23 per user per month, or $18 billed annually) and Enterprise lock training off for every member; Dictation Cloud Storage can still remain individually controllable, so "check both." - **For healthcare in the US:** a signed BAA locks both controls off and blocks Notetaker recording. ## What Wispr Flow does well on privacy Wispr Flow's privacy documentation is more detailed than most US dictation tools, and several defaults are genuinely good. A fair assessment has to say so. - **Clear separation of controls:** training, storage and local history are documented as independent switches. - **Organisation defaults:** training is off and locked for Growth and Enterprise customers. - **Fail-safe:** when Flow cannot confirm your preference, it does not share dictation audio. - **Provider contracts:** zero-retention and no-training agreements with third-party AI providers for dictation. - **Data export:** a self-service ZIP export of dictations, meetings, notes and settings, available for 30 days. - **US compliance:** SOC 2 Type II, ISO/IEC 27001:2022 and HIPAA support with a signed BAA. ## How Sprecho handles retention Sprecho puts retention behind one setting that is enforced on the server before anything is written, and all processing happens in the EU. In the Sprecho desktop app you open Settings → Privacy → Data Retention and set Text Retention and Audio Retention to **"Delete immediately after pasting"**. - **Server-side enforcement:** with text set to delete immediately, "the transcript row is never created"; the server returns the formatted text and writes nothing, so sync has nothing to send to other devices. - **Audio:** with audio set to delete immediately, the app keeps no local copy and the uploaded file is deleted as soon as processing finishes. - **Back catalogue purged:** switching text to "delete immediately" also purges every transcript already stored; audio is cleared with Delete All Data. - **Honest defaults:** out of the box Sprecho keeps audio for 14 days and text until you delete it; you choose between immediate deletion, fixed windows from 1 day to 3 months, or keeping text. - **Training:** customer audio and transcripts are never used to train Sprecho's models. A separate "Help improve Sprecho" switch, on unless you turn it off, contributes anonymised text patterns only, never audio, with names and addresses replaced on your device. - **Team policy:** a team owner's retention policy is enforced on the server for every member, and the stricter setting wins. - **EU infrastructure:** storage at STRATO GmbH in Germany, GPU inference at Media Trooper in Germany and the Netherlands, no third-party AI provider, no US sub-processor in the dictation path. | Question | Wispr Flow | Sprecho | |---|---|---| | **Training on customer dictations** | Possible if "Improve the model for everyone" is on (default on Free and Pro) | ❌ Never; optional anonymised text patterns only, no audio | | **One setting for "store nothing"** | ❌ Two switches plus Notetaker, Scratchpad and feedback caveats | ✅ Text and Audio Retention: "Delete immediately after pasting" | | **Enforced on the server** | Organisation policies and BAA locks enforced; settings reset to cloud storage on after sign-out | ✅ Retention resolved before any row is written | | **Default retention** | Dictation Cloud Storage reported on by default | Audio 14 days, text kept until deleted (both adjustable) | | **Meeting recorder storage** | Notetaker transcripts stored, no opt-out; audio 7 days | No meeting recorder; dictation only | | **Where processing happens** | United States | ✅ Germany and the Netherlands | | **Company jurisdiction** | US (Delaware); CLOUD Act applies | ✅ Germany (Melo Designer GmbH) | | **Third-party AI providers** | Yes, under zero-retention agreements for dictation | ✅ None; inference on Sprecho-operated infrastructure | | **Price for one user (Pro)** | $15 per month or $12 per month billed annually (shown as €15 / €12 to German visitors; no VAT added, reverse charge for businesses) | €10.92 net per month or €9.24 net per month yearly (€12.99 / €10.99 incl. 19 % VAT) | ![Quiet office at dusk with a closed laptop under a desk lamp, symbolising data that is no longer stored](https://api.sprecho.ai/media/2026/10/1791204608-a6f499fa.jpg) Sprecho is cloud-based too and has no offline mode; the difference lies in where the cloud is, who operates it, and whether "store nothing" is one enforced setting. The step-by-step guide is in our help article on [turning off transcript and audio storage](https://sprecho.ai/help/privacy-and-security/turn-off-transcript-and-audio-storage), and the full inventory sits in [what Sprecho stores, and where](https://sprecho.ai/help/privacy-and-security/what-sprecho-stores-and-where). ## Who should care most about these settings Anyone who dictates personal data should check both Wispr switches; professionals bound by secrecy duties need more than settings. The right answer depends on your role. ### Lawyers and tax advisors - **§ 203 StGB:** disclosing client secrets to a service provider is only permitted under the conditions of § 203 (3) and (4) StGB and, for lawyers, § 43e BRAO, which includes a written confidentiality obligation. - **Practical consequence:** a training switch does not replace a contract with a provider you can bind under German law; see [Sprecho for law firms](https://sprecho.ai/solutions/legal) and [Sprecho for tax advisors](https://sprecho.ai/solutions/tax-consultant). ### Doctors and healthcare teams - **GDPR Art. 9:** health data is a special category; dictating findings into a US cloud needs a solid legal basis and transfer assessment. - **HIPAA is US law:** a BAA addresses US healthcare rules, not German professional secrecy; see [Sprecho for healthcare](https://sprecho.ai/solutions/healthcare). ### IT departments and data protection officers - **Art. 28 GDPR:** you need a data processing agreement, the full sub-processor list and the TOMs under Art. 32. - **Transfers:** after Schrems II (CJEU C-311/18, 16 July 2020), US transfers need a transfer impact assessment; read our explainer [where is my voice stored in AI dictation](https://sprecho.ai/blog/where-is-my-voice-stored-ai-dictation) and the help article on [GDPR and your data processing agreement](https://sprecho.ai/help/privacy-and-security/gdpr-and-your-data-processing-agreement). ## Conclusion Wispr Flow's former Privacy Mode is a training switch, not a retention switch: it keeps your dictations out of Wispr's model training, but your transcripts, audio and history stay on US servers until you also turn off Dictation Cloud Storage, and Notetaker transcripts, Scratchpad notes, dictionaries, feedback and usage statistics remain outside both switches. Wispr documents all of this openly, and for English-first users without secrecy duties a carefully configured Wispr Flow can be a reasonable choice. For German professionals who need "store nothing" to be one enforced setting, processed in the EU by an EU company, **Sprecho offers server-enforced "Delete immediately after pasting" with EU processing and a public AVV, and you can test it free for 14 days without a card**. Compare both tools in detail on [Sprecho vs Wispr Flow](https://sprecho.ai/compare/sprecho-vs-wispr-flow), see the broader field in [the best Wispr Flow alternatives for Germany](https://sprecho.ai/blog/wispr-flow-alternatives-germany), or check [Sprecho pricing](https://sprecho.ai/pricing). ## FAQ ### What does Wispr Flow's Privacy Mode do? Privacy Mode has been renamed "Improve the model for everyone" and only controls whether Wispr may use your audio, transcripts and edits to train or evaluate its own AI models. Wispr states that if you had Privacy Mode enabled, your data is still not shared for model training. It does not control whether your dictations are stored on Wispr's servers. ### Does Wispr Flow store my audio? It can. Dictation audio, transcripts and history are stored on Wispr's servers in the United States while Dictation Cloud Storage is on, and transcription always happens in the cloud. Notetaker meeting audio is kept for 7 days from upload, then deleted, while meeting transcripts and summaries remain. ### What is Dictation Cloud Storage in Wispr Flow? Dictation Cloud Storage is the setting that decides whether Wispr stores your transcription data, meaning transcripts, audio and dictation history, on its servers. It covers dictation only, not Notetaker or Scratchpad. Wispr's help center notes that signing out resets it to on, so check it after every sign-in. ### How do I get zero data retention in Wispr Flow? Turn off both "Improve the model for everyone" and Dictation Cloud Storage in Settings → Data and Privacy, then avoid Notetaker and feedback on sensitive content. Even then, account data, dictionaries, snippets, desktop Scratchpad notes and usage statistics such as word counts remain stored. Audio is still processed in Wispr's US cloud for every dictation. ### Does Wispr Flow's Context Awareness take screenshots? According to Wispr's Data Controls page, Context Awareness uses "relevant text data from the active app window" to improve accuracy, not screenshots. Independently of that switch, Wispr always uses the app name and the surrounding textbox content for formatting and Command Mode. You can turn Context Awareness off in Settings → Data and Privacy at any time. ### Does Wispr Flow train on my data? Wispr may use your audio, transcripts and edits to train its own models if "Improve the model for everyone" is on, which is the default on the Free and Pro plans. On Growth and Enterprise the setting is off and cannot be re-enabled by administrators. Wispr says its third-party AI providers cannot use your data for training in any case. ### Can I delete my data in Wispr Flow? Yes. You can request a ZIP export of your stored dictations, meetings, notes and settings, available for 30 days, and request account deletion in Settings → Data and Privacy. Wispr warns that deleted records cannot be restored; contractual retention periods and deletion confirmations are handled by your account representative. ### How does Sprecho handle data retention? In Sprecho you set Text Retention and Audio Retention to "Delete immediately after pasting" under Settings → Privacy → Data Retention, and the server then writes no transcript at all. Defaults are 14 days for audio and keep-until-deleted for text, adjustable from immediate deletion to 3 months. All processing runs in Germany and the Netherlands under Melo Designer GmbH, with an AVV under Art. 28 GDPR. --- ## https://sprecho.ai/blog/dragon-legal-anywhere-vs-sprecho-german-law-firms-2026 — Dragon Legal Anywhere vs Sprecho for German Law Firms: BRAO, § 203 and Who Hosts Your Dictation (2026) > Dragon Legal Anywhere vs Sprecho: who hosts your dictation, BRAO and § 203 StGB duties, Mac and mobile support, and prices for a five-lawyer firm in 2026. *Published: 2026-11-20T18:03:00.473000Z* *Tags: dragon, legal, gdpr, comparison, voice-dictation, germany, pricing* **Dragon Legal Anywhere** is the cloud edition of Nuance’s legal speech recognition, and for many German law firms it is the default answer to “how do we dictate Schriftsätze without typing?”. In 2026 the more interesting question is a different one: **who hosts your dictation, which company signs the contract, and does that chain survive a § 203 StGB and BRAO review?** This comparison puts Dragon Legal Anywhere and Sprecho side by side on exactly those points, plus platforms, features and price for a typical five-lawyer Kanzlei. *Last verified: October 2026.* We are the team behind Sprecho, so read this with that in mind. We have tried to be fair: where Dragon Legal is stronger, we say so plainly, and every Dragon fact below is sourced from Nuance’s own pages or named reseller listings. ## Dragon Legal Anywhere vs Sprecho at a glance In short: **Dragon Legal Anywhere is a Windows-client legal cloud with decades of German legal vocabulary; Sprecho is a cross-platform EU dictation service run entirely by German and EU companies.** Both are subscriptions, both are sold as GDPR-compliant, but their vendor chains and platform reach differ sharply. ### Dragon Legal Anywhere: the Nuance legal cloud - **What it is:** a cloud-hosted speech recognition service for lawyers, notaries and legal departments, with a legal language model and a lightweight Windows client that streams audio to Nuance’s servers. - **Vendor:** Nuance Communications, part of Microsoft since the acquisition closed on 4 March 2022. In Germany it is sold through resellers. - **Marketing claims:** [Nuance’s German product page](https://dragon.nuance.com/de-de/dragon-legal-anywhere) promises **99 % recognition accuracy** for legal terminology, dictation **“three times faster than typing”**, 256-bit encryption in transit and at rest, and storage that meets DSGVO, BDSG (neu) and ISO 27001. - **Client platforms:** the [Dragon Legal Anywhere 2026.1 release notes](https://supportcontent.nuance.com/kbfiles/DLA_RELEASE_NOTES_2026.1-en.htm) list Windows 10 and Windows 11 as client operating systems, plus Citrix, VMware Horizon and Remote Desktop environments. ### Sprecho: EU dictation from Melo Designer GmbH - **What it is:** a push-to-talk dictation app that writes into any application (Word, Outlook, RA-MICRO, DATEV, beA web forms) with AI formatting, a personal and team dictionary, and spoken snippets. - **Vendor:** Melo Designer GmbH in Lastrup, Germany, with a TÜV SÜD-certified data protection officer. The AVV (DPA under Art. 28 GDPR) is signed by the German GmbH. - **Infrastructure:** STRATO GmbH servers in Germany, GPU inference at Media Trooper (Germany/Netherlands), off-site backups at Hetzner in Nürnberg. **No US sub-processor in the dictation path.** - **Platforms:** native apps for Windows, macOS and Linux plus iOS and Android, three devices per licence. ## What is Dragon Legal Anywhere, and what changed in 2026? **Dragon Legal Anywhere is Nuance’s subscription cloud for legal dictation, and in October 2026 it is still sold and updated, but authorised Nuance partners have announced its end of sale for 31 December 2026:** most of the rest of the Dragon line has already been discontinued or shifted to Microsoft infrastructure. - **Release cadence:** cloud versions are retired on a rolling basis. According to the 2026.1 release notes, **Dragon Legal Anywhere 2024.2 and all earlier versions reached end of life on 31 August 2026**, version 2024.3 on 30 September 2026, and version 2025.1 is scheduled for 13 February 2027. IT has to keep the client current. - **Sibling product:** Dragon Legal 16 is the perpetual Windows desktop edition from the version 16 generation that Nuance launched in February 2023. There is no version 17 as of October 2026. - **Mobile:** Nuance’s German page still mentions the Dragon Anywhere app as the mobile companion, but its [App Store listing](https://apps.apple.com/us/app/dragon-anywhere/id1024652126) states that since **1 July 2026** new subscriptions and renewals are no longer possible. - **Mac:** Dragon for Mac was discontinued on 22 October 2018 ([Nuance support notice](https://nuance.custhelp.com/app/answers/detail/a_id/27843/)). There is no native Mac client for any Dragon Legal product. - **Where Dragon Legal remains strong:** a legal language model refined over decades, automatic formatting of legal citations, and a mature voice-command grammar for selecting and correcting text by voice. ### Partner-announced end of sale: 31 December 2026 - **The announcement:** two authorised Nuance partners in Germany, Thax Software (3 September 2026) and Byte Heroes (15 July 2026), state that Dragon Professional Anywhere and Dragon Legal Anywhere reach **end of sale on 31 December 2026** and **end of life on 31 December 2027**. - **Existing contracts:** according to both partners, contracts can be renewed one last time in 2026 and then run out in 2027. This applies to all contract forms: Azure cloud, Telekom-hosted and on-premise with maintenance. - **No direct successor:** Thax writes that Nuance has no direct successor for the Anywhere editions and names Dragon Professional 16, Dragon Legal 16 or Dragon Medical One as alternatives. - **Status of the source:** as of October 2026 there is no public Nuance or Microsoft advisory; Nuance’s release notes only publish per-version lifecycles. Treat the product-level dates as partner-announced and **get them confirmed in writing** by your reseller. ## Who hosts your dictation? Telekom, Azure and the vendor chain **For German customers, Dragon Legal Anywhere has been hosted in the Open Telekom Cloud of Deutsche Telekom, which is a genuinely better hosting story than a US hyperscaler.** But hosting is only one link in the chain; the software vendor and contract partner matter too, and they are not the same company. ### Layer 1: the hosting provider - **Open Telekom Cloud:** German Nuance resellers describe Dragon Legal Anywhere as hosted in the Open Telekom Cloud of Deutsche Telekom, a German company operating data centres in Germany. That is the setup most German Kanzleien have bought. - **Azure is also in play:** at least one German reseller now lists Microsoft Azure (“Nuance Cloud”) as an additional hosting option, and Nuance’s own English data sheet names Microsoft Azure as the hosting infrastructure for the service. - **What Nuance itself says in German:** the German product page promises DSGVO-, BDSG- and ISO 27001-compliant storage but **does not name the hosting provider** (checked October 2026). - **Practical consequence:** ask your reseller in writing which platform your tenant runs on, and get it into the contract. Do not assume. ### Layer 2: the vendor and the contract partner - **The software vendor is Nuance:** Nuance develops, operates and updates the service, regardless of whose data centre it runs in. Since March 2022, Nuance is a subsidiary of Microsoft Corporation, a US company. - **The contracting entity:** Nuance’s published [Dragon Anywhere terms](https://dragon.nuance.com/en-us/dragon-anywhere-eula.html) name Nuance Communications Ireland Limited as the contracting party for customers outside North America and Japan. Check your Dragon Legal Anywhere order form and DPA for the exact entity in your case. - **The reseller:** in Germany you usually buy through a reseller, who may add support and training but is not the processor operating the recogniser. - **Sub-processors:** Deutsche Telekom (or Microsoft, for Azure tenants) appears as the infrastructure sub-processor of Nuance. ### What the CLOUD Act does and does not mean here - **The rule:** the US CLOUD Act (2018) lets US authorities compel providers subject to US jurisdiction to disclose data in their “possession, custody, or control”, wherever the data is stored. - **Telekom-hosted data is not on a US cloud:** Deutsche Telekom is not a US company, so the hosting provider itself is not a CLOUD Act addressee. That is a real advantage over Azure-hosted products such as Dragon Medical One. - **The open question is control:** whether a Microsoft-owned vendor that operates the service has “control” over data in a third-party German cloud is a legal assessment your firm (or its DPO) has to make. The exposure is **lower and indirect**, but a US group sits in the chain by design. - **Sprecho’s chain has no US entity:** contract partner, hosting and GPU inference are German or Dutch companies, so there is no US parent that could be compelled in the first place. | Product | Vendor / contract partner | Where recognition runs | Infrastructure owner | CLOUD Act exposure | |---|---|---|---|---| | Dragon Legal 16 (desktop) | Nuance (Microsoft group) | On your Windows PC | Your firm | ✅ None for dictation data (stays local) | | Dragon Legal Anywhere (Telekom tenant) | Nuance (Microsoft group), sold via resellers | Open Telekom Cloud, Germany | Deutsche Telekom (DE) | Indirect: German host, US-owned vendor; assess “control” | | Dragon Legal Anywhere (Azure tenant) | Nuance (Microsoft group) | Microsoft Azure | Microsoft (US) | ❌ Yes, US hyperscaler | | Dragon Medical One (for context) | Nuance (Microsoft group) | Azure Frankfurt / Berlin | Microsoft (US) | ❌ Yes, despite German data centres | | Sprecho | Melo Designer GmbH (DE) | STRATO (DE) + Media Trooper (DE/NL) | German / EU companies | ✅ No US entity in the dictation path | ## BRAO and § 203 StGB: what your Kanzlei has to check **Both tools can be used lawfully by German lawyers, provided the outsourcing meets § 43e BRAO and § 203 StGB; the difference is how much work and residual risk that review leaves you with.** No vendor can “be BRAO-compliant” on your behalf; the duty sits with the lawyer. - **§ 43e BRAO, paragraph 1:** you may give a service provider access to privileged facts only as far as the service requires it. - **§ 43e BRAO, paragraph 3:** the contract must be in text form, oblige the provider to confidentiality with a warning about the criminal consequences, limit its knowledge to what the job needs, and regulate whether it may involve third parties. - **§ 43e BRAO, paragraph 4:** if the service is performed abroad, you may open access only if the protection of secrets there is **comparable to protection in Germany**. A US parent company is exactly where that question gets uncomfortable. - **§ 203 StGB:** paragraph 3 allows disclosure to “other persons involved” where necessary; paragraph 4 makes you liable if you did not ensure they were bound to secrecy. The penalty is up to one year’s imprisonment or a fine. - **GDPR Art. 28 and Art. 9:** you need a processing agreement (AVV), and if a mandate concerns health data (personal injury, medical liability, social law), Art. 9 GDPR applies to the dictated content. Nuance is candid that this analysis is the customer’s job. Its [Dragon Anywhere terms](https://dragon.nuance.com/en-us/dragon-anywhere-eula.html) state that professionals bound by secrecy laws “are solely responsible to analyze whether the conditions set out in this Agreement, the DPA, and the way that we process the data … are adequate to preserve your legal and deontological obligations”. The German Dragon Legal Anywhere page does not mention BRAO or § 203 at all (checked October 2026). ![Reception of a German law firm with case files and a workstation](https://api.sprecho.ai/media/2026/10/1791196904-4fd4d91f.jpg) ### A practical BRAO checklist for any dictation vendor 1. **Name the processor:** which legal entity operates the recogniser, and which group does it belong to? 2. **Name every sub-processor:** hosting, GPU inference, backups, support tooling. 3. **Get the § 43e undertaking:** a confidentiality obligation in text form with the criminal-law warning. 4. **Clarify retention:** how long are audio and text kept, and can you switch storage off? 5. **Check third-country access:** is any company in the chain subject to US law? For Sprecho the answers fit on one page: **Melo Designer GmbH** as processor, STRATO, Media Trooper and Hetzner as named sub-processors, an AVV for Enterprise customers, and retention you can set to **“delete immediately after pasting”** for both transcripts and audio, enforced on the server. More detail is on our [Sprecho for law firms page](https://sprecho.ai/solutions/legal). ## Feature comparison: Dragon Legal Anywhere vs Sprecho **Dragon Legal Anywhere wins on legal vocabulary depth and voice-command grammar; Sprecho wins on platforms, AI formatting and vendor chain.** The table shows both honestly. | Feature | Dragon Legal Anywhere | Sprecho | |---|---|---| | Windows client | ✅ Windows 10/11, Citrix, RDS | ✅ Native app | | macOS | ❌ No Mac client | ✅ Native app | | Linux | ❌ | ✅ Native app | | iOS / Android | ❌ Dragon Anywhere off sale since 1 July 2026 | ✅ Native apps, same account | | Dictates into any application | ✅ Windows applications | ✅ Any app on every platform | | Legal vocabulary out of the box | ✅ Decades-old legal model, citation formatting | ✅ Fachsprache dictionary plus your own entries | | Custom words and replacements | ✅ Vocabulary editor | ✅ Dictionary, misspelling rules, auto-learn from corrections | | Firm-wide shared vocabulary | ✅ Via Nuance management tools | ✅ Shared team dictionary (Pro / Enterprise) | | Boilerplate by voice | ✅ Auto-texts | ✅ Snippets with {date}, {time}, {clipboard} | | Voice command grammar (select, correct, cap) | ✅ Very extensive | ❌ High-frequency set only (punctuation, new line, “Streich das”) | | AI formatting and per-app style | ❌ | ✅ Formatting models per app category | | Languages | Legal edition per language | ✅ 100+ with automatic detection | | Vendor jurisdiction | ❌ Microsoft group (US parent) | ✅ German GmbH | | Storage switch-off by user | Not documented publicly | ✅ “Delete immediately” for text and audio | | SSO / SAML, MDM rollout | Enterprise tooling via Nuance | ✅ Enterprise plan | ### Where Dragon Legal is stronger - **Legal language model:** Nuance has trained on legal text for decades, and the citation formatting is genuinely mature. - **Command grammar:** “select”, “correct”, “cap that”, inserting before and after; if you edit entire documents hands-free, nothing else comes close. - **Virtual desktop support:** documented Citrix, VMware Horizon and Remote Desktop support for terminal-server Kanzleien. - **Offline option in the family:** Dragon Legal 16 runs fully on-device, which no cloud service, Sprecho included, can match. ### Where Sprecho is stronger - **Every device in the firm:** Windows, Mac, Linux, iPhone and Android with one account, one dictionary and the same snippets. - **Less command overhead:** AI formatting adds punctuation, removes filler words and structures paragraphs, so you dictate content instead of commands. - **A shorter vendor chain:** one German GmbH as processor, no US group anywhere in the dictation path. - **Transparent pricing:** list prices on the website, a free plan and a 14-day Pro trial without a card. ## Mac partners, mobile dictation and the Windows-only client **If even one partner works on a MacBook or dictates on the phone, Dragon Legal Anywhere leaves a gap that the firm has to fill with a second tool.** That is the most common reason Kanzleien contact us. ![Law firm partner working on a MacBook in a meeting room](https://api.sprecho.ai/media/2026/10/1791196904-a0dbe0b9.jpg) - **MacBook partners:** there is no Mac client. Running Windows in a virtual machine only lets you dictate into Windows applications inside the VM, never into native Mac Mail or Pages. - **Phone dictation:** with Dragon Anywhere off sale since 1 July 2026, Nuance has no mobile dictation product you can buy or renew. - **Two tools, two vocabularies:** a second app for Mac or mobile means a second dictionary that drifts apart from the Windows one. - **Sprecho’s model:** the same app on every operating system, three devices per licence, and a dictionary that follows the user. ## How Sprecho fits the daily work of a Kanzlei **Sprecho works wherever the cursor is: you hold the hotkey, speak, and formatted text lands in Word, RA-MICRO, DATEV, Outlook or a beA web form.** There is no plugin per application. - **Schriftsätze:** dictate the Sachverhalt in Word; AI formatting sets punctuation and paragraphs, and the voice command “section sign” (in German “Paragraphenzeichen”) inserts § exactly where you need it. - **Aktenzeichen and names:** put case numbers, parties and opposing counsel into the [dictionary](https://sprecho.ai/help/formatting-and-ai/teach-sprecho-your-words-with-the-dictionary); misspelling rules fix recurring errors automatically. - **Boilerplate:** turn “Namens und in Vollmacht meiner Mandantschaft …” or your Beweisantritt into [snippets](https://sprecho.ai/help/formatting-and-ai/snippets-voice-shortcuts-for-text-you-repeat) with a spoken trigger. - **Different tone per app:** with [per-app style](https://sprecho.ai/help/formatting-and-ai/set-a-writing-style-per-app), emails can be formatted more fully than quick Teams messages. - **Hold or lock mode:** hold the key for short notes, lock it for a long dictation of several pages. ### For the firm, not just one lawyer - **Shared team dictionary and snippets:** one firm-wide vocabulary on Pro and Enterprise, combined with each lawyer’s personal entries. - **Enterprise administration:** admin dashboard, licence management, SSO/SAML, MDM deployment via Intune or Jamf, audit-log export and payment by invoice. - **AVV from a German GmbH:** signed by Melo Designer GmbH, with a published sub-processor list. - **Migration help:** export your Dragon word list and send it to our support team; we convert it into dictionary entries. ## Pricing: Dragon Legal Anywhere vs Dragon Legal 16 vs Sprecho **For five lawyers over one year, Sprecho Pro costs about 554 € net, while Dragon Legal Anywhere costs roughly 2,894 € to 3,216 € net at regular yearly reseller prices.** Dragon prices in Germany are reseller quotes and vary, so treat the figures as observed ranges, not list prices. | Option | Pricing model | Price per seat (net) | 5 lawyers, first year (net) | |---|---|---|---| | Dragon Legal Anywhere (STP marketplace) | Yearly subscription | €643.20/yr regular (€298.80 first year with promo code) | €3,216.00 (regular price; €1,494.00 with first-year promo) | | Dragon Legal Anywhere (software-express.de) | Yearly subscription | €578.77/yr | €2,893.85 | | Dragon Legal Anywhere (other resellers) | Yearly subscription | €619.00/yr (previously €703.80) | €3,095.00 | | Dragon Legal Anywhere (softguide.de listing) | Monthly, 24-month minimum term | €69.00/mo | €4,140.00 | | Dragon Legal 16 (perpetual, Windows) | One-time licence | €1,007.56 (1–9 seats); upgrade from v15 €449 | €5,037.80 one-time | | Sprecho Pro | Monthly or yearly | €10.92/mo or €9.24/mo yearly (€110.82/yr) | €554.10 (yearly) | | Sprecho Enterprise | Monthly or yearly | €21.84/mo or €18.47/mo yearly (€221.64/yr) | €1,108.20 (yearly) | | Sprecho On-premise | On request | On request | Customer-hosted deployments (~50+ seats) | - **Gross prices:** Sprecho Pro is 12.99 € per month or 10.99 € per month billed yearly, Enterprise 25.99 € or 21.99 €, all including 19 % VAT. - **Free start:** Free Forever with 2,000 words per week and a 14-day Pro trial, no card required. Enterprise tests run for 14 days. - **Dragon trials:** German resellers offer free trial periods for Dragon Legal Anywhere; ask for the length in writing. - **Mind the announced end of life:** a Dragon Legal Anywhere subscription signed or renewed now runs into the partner-announced end of life on 31 December 2027, so a five-lawyer firm budgets for a second migration within roughly a year. - **Hidden costs to include:** Windows VMs for Mac partners, a separate mobile tool, and client updates whenever a cloud version reaches end of life. All Sprecho plans and the current prices are on the [pricing page](https://sprecho.ai/pricing). ## Moving from Dragon Legal to Sprecho **A five-lawyer firm can switch in an afternoon, because nothing needs acoustic training and the vocabulary can come with you.** We recommend running both tools in parallel for the 14-day trial. 1. **Export your vocabulary:** in Dragon desktop editions, open the Vocabulary Center and choose “Export custom word and phrase list” (TXT or XML); for Legal Anywhere, ask your Dragon administrator for the export. 2. **Send it to us:** there is no self-service bulk import yet; send us your exported list and we convert it into dictionary entries. 3. **Recreate auto-texts as snippets:** letterheads, Beweisantritte, closing formulas. 4. **Set your hotkey:** choose the key you used for Dragon and decide between hold and lock mode. 5. **Share with the team:** move firm-wide terms into the shared team dictionary and sign the AVV. ## Other tools law firms ask about **Three other names come up regularly in Kanzlei evaluations; none of them is a like-for-like replacement for Dragon Legal.** - **Wispr Flow:** fast, polished flow dictation from a US company; check jurisdiction carefully. See [Sprecho vs Wispr Flow](https://sprecho.ai/compare/sprecho-vs-wispr-flow). - **Otter.ai:** a meeting transcription tool from the US, not system-wide dictation. See [Sprecho vs Otter.ai](https://sprecho.ai/compare/sprecho-vs-otter-ai). - **Apple Dictation:** free and built in, but without a legal dictionary or team features. See [Sprecho vs Apple Dictation](https://sprecho.ai/compare/sprecho-vs-apple-dictation). ## Which one should your firm choose? **Choose by platforms and vendor chain first, vocabulary second.** - **Stay with Dragon Legal Anywhere if:** every lawyer works on Windows or Citrix, you edit long documents purely by voice command, and your DPO has signed off on the Nuance chain for your tenant. - **Consider Dragon Legal 16 if:** you need fully offline recognition on Windows and accept a perpetual licence with no newer major version since 2023. - **Choose Sprecho if:** partners use Macs or phones, you want one German GmbH as processor without a US parent, and you prefer AI formatting over command grammar. - **Test before you decide:** dictate the same Schriftsatz in both tools for a week and compare corrections per page. ## Conclusion Dragon Legal Anywhere is a serious legal product, and its Telekom hosting makes it the most privacy-conscious of Nuance’s cloud offerings in Germany. Its limits are structural: partners have announced its end of sale for 31 December 2026 and end of life for 31 December 2027, a Windows-only client, no mobile product since July 2026, reseller pricing of roughly 579 € to 643 € net per seat and year (first-year promotions from 298.80 €), and a Microsoft-owned vendor in the chain even when the servers belong to Deutsche Telekom. **If your Kanzlei wants one dictation tool on every device, run by a German GmbH with no US entity in the path, start a free 14-day Sprecho trial today** and read our [full Sprecho vs Dragon comparison](https://sprecho.ai/compare/sprecho-vs-dragon) or the overview of the [best Dragon alternatives for 2026](https://sprecho.ai/blog/best-dragon-alternatives-2026). ## FAQ ### Where is Dragon Legal Anywhere hosted? German Nuance resellers state that Dragon Legal Anywhere is hosted in the Open Telekom Cloud of Deutsche Telekom in Germany, and at least one now lists Microsoft Azure as an additional option. Nuance’s German product page promises DSGVO-, BDSG- and ISO 27001-compliant storage but does not name the provider. Ask your reseller in writing which platform your tenant runs on. ### Is Dragon Legal Anywhere BRAO-compliant? No vendor can be BRAO-compliant on a lawyer’s behalf: § 43e BRAO puts the duty on the lawyer to choose the provider carefully, sign a confidentiality agreement in text form and, for services abroad, check that protection is comparable to Germany. Dragon Legal Anywhere can be used within that framework, but your review has to cover Nuance as a Microsoft subsidiary, not only the Telekom data centre. ### Does Dragon Legal Anywhere run on a Mac? No. The client runs on Windows 10 and Windows 11 and in virtual desktop environments such as Citrix; there is no Mac client, and Dragon for Mac was discontinued in October 2018. Sprecho has native apps for macOS, Windows and Linux plus iOS and Android. ### What does Dragon Legal Anywhere cost in Germany? Prices are reseller quotes and vary: regular yearly prices range from about 579 € net (software-express.de) to 643.20 € net (STP marketplace), with 619 € net at other resellers and first-year promotions as low as 298.80 € net; softguide.de lists 69 € per month net with a 24-month term. Sprecho Pro costs 9.24 € net per month billed yearly (110.82 € per year), and Enterprise 18.47 € net per month billed yearly. ### Is Sprecho suitable for lawyers under § 203 StGB? Sprecho is built for that review: the processor is Melo Designer GmbH in Germany, servers are at STRATO in Germany, GPU inference at Media Trooper in Germany and the Netherlands, and there is no US company in the dictation path. Enterprise customers receive an AVV, and storage of transcripts and audio can be switched off so that nothing is retained after pasting. ### Does Sprecho work in RA-MICRO, DATEV and beA? Yes. Sprecho dictates system-wide into whatever field has the cursor, so it works in RA-MICRO, DATEV, Word, Outlook and beA web forms without a plugin. You hold the hotkey, speak, and the formatted text is inserted. ### Can a five-lawyer firm share one dictionary in Sprecho? Yes. On Pro and Enterprise a team-shared dictionary and shared snippets are available, and they combine with each lawyer’s personal entries. If you are coming from Dragon, send us your exported word list and we convert it into dictionary entries. ### Is dictated voice data biometric data under GDPR Art. 9? Voice recordings become biometric data when they are processed to identify a speaker; dictation for transcription usually does not do that. Art. 9 GDPR still matters for law firms because dictated content often contains health data, for example in personal injury or medical liability mandates, which is why the processor, its location and its retention settings deserve a careful review. --- ## https://sprecho.ai/blog/is-dragon-discontinued-status-of-every-dragon-product-2026 — Is Dragon Discontinued? The Status of Every Dragon Product in October 2026 > Dragon status October 2026: what is discontinued, what is still sold, with dates and official sources for Dragon 16, Anywhere, Medical One and Copilot. *Published: 2026-11-03T18:27:06.444000Z* *Tags: dragon, voice-dictation, gdpr, germany, windows, mac, mobile, migration* **Is Dragon discontinued?** Not all of it, but most of what consumers, Mac users and mobile users bought is gone: Dragon for Mac ended in 2018, Dragon Home in 2023 and Dragon Anywhere on 1 July 2026, while Dragon Professional 16 for Windows and the cloud editions for medicine and law are still sold. This page lists the status of every Dragon product with a date and an official source, explains what each status means in practice, and shows where German professionals can go next. *Last verified: October 2026.* Dragon is the speech recognition family that Nuance Communications built over three decades and that Microsoft has owned since it completed the Nuance acquisition on 4 March 2022 (about $19.7 billion). We update this page quarterly; if a date changes, the table changes with it. ## Is Dragon discontinued? The short answer **Dragon as a brand is not discontinued, but the product range has shrunk to Windows desktop licences and Microsoft-hosted cloud subscriptions.** Everything outside Windows has ended, and the general-purpose desktop line has not had a major release since February 2023. - **Discontinued:** Dragon Professional Individual for Mac (22 October 2018), Dragon Home (last sold as version 15 until 27 February 2023) and the Dragon Anywhere mobile app (no sales or renewals since 1 July 2026). - **Still sold:** Dragon Professional 16 and Dragon Legal 16 for Windows, Dragon Medical One and Dragon Copilot for healthcare, plus the cloud subscriptions Dragon Professional Anywhere and Dragon Legal Anywhere. - **Announced by partners, not yet by Nuance:** authorised Nuance partners in Germany report an end of sale for Dragon Professional Anywhere and Dragon Legal Anywhere on 31 December 2026 and an end of life on 31 December 2027. - **No Dragon 17:** as of October 2026 Nuance has not announced a successor to Dragon Professional 16, which was released on 28 February 2023. - **Where the money goes:** Microsoft's visible investment is in clinical AI (Dragon Copilot), not in the general-purpose desktop product. ## The status of every Dragon product (October 2026) **The master table below gives each Dragon product a status, a date and the source that date comes from.** Where Nuance has not published a date, the table says "not published" instead of guessing. | Product | Type | Status | Date | Source | |---|---|---|---|---| | **Dragon NaturallySpeaking** | Windows desktop | Name retired; line continues as Dragon Professional | Version 13 (July 2014) was the last major release under that name | [Nuance advisory (successor editions)](https://nuance.custhelp.com/app/answers/detail/a_id/29838/) | | **Dragon Home** | Windows desktop (consumer) | ❌ Discontinued, no successor | Dragon Home 15 sales ended 27 Feb 2023 | [Nuance advisory 29838](https://nuance.custhelp.com/app/answers/detail/a_id/29838/) | | **Dragon Professional Individual for Mac 6** | macOS desktop | ❌ Discontinued | 22 Oct 2018 | [Nuance advisory 27843](https://nuance.custhelp.com/app/answers/detail/a_id/27843/) | | **Dragon Anywhere** | iOS / Android app | ❌ No new subscriptions, no renewals | 1 Jul 2026 (last app version 1.96.1) | [App Store listing](https://apps.apple.com/us/app/dragon-anywhere/id1024652126) | | **Dragon Professional 14** | Windows desktop | ❌ All support ended | 1 Sep 2023 | [Nuance advisory 29838](https://nuance.custhelp.com/app/answers/detail/a_id/29838/) | | **Dragon Professional 15** (incl. Legal 15, Law Enforcement 15) | Windows desktop | ❌ No longer sold (EN + DE) | Sales ended 27 Feb 2023 | [Nuance advisory 29838](https://nuance.custhelp.com/app/answers/detail/a_id/29838/) | | **Dragon Professional 16** | Windows desktop, perpetual | ✅ Sold and supported | Released 28 Feb 2023; Dragon 17: not published | [Nuance advisory 29838](https://nuance.custhelp.com/app/answers/detail/a_id/29838/) | | **Dragon Legal 16** | Windows desktop, perpetual (USA + Germany) | ✅ Sold | v16 generation since 28 Feb 2023; end date not published | [Nuance advisory 29838](https://nuance.custhelp.com/app/answers/detail/a_id/29838/) | | **Dragon Professional Anywhere** | Cloud subscription (Microsoft Azure) | ✅ Sold; end of sale announced by partners | Partners: end of sale 31 Dec 2026, end of life 31 Dec 2027; release 2024.2 EOL 31 Aug 2026 | [Release notes 2026.1](https://supportcontent.nuance.com/kbfiles/DPA_RELEASE_NOTES_2026.1-en.htm) | | **Dragon Legal Anywhere** | Cloud subscription (Deutsche Telekom Open Telecom Cloud) | ✅ Sold; end of sale announced by partners | Partners: end of sale 31 Dec 2026, end of life 31 Dec 2027; release 2024.1 EOL 20 Feb 2026 | [Release notes 2024.2](https://supportcontent.nuance.com/kbfiles/DLA_RELEASE_NOTES_2024.2.htm) | | **Dragon Medical One** | Cloud subscription (Azure, German Hosted) | ✅ Sold; being folded into Dragon Copilot | Rolling release EOL; LTSR 2021.4.5 and older: after 31 Dec 2026 (integrator report) | [Microsoft newsroom](https://news.microsoft.com/de-de/nach-erfolgreicher-pilotphase-microsoft-dragon-copilot-fuer-deutschlands-kliniken-und-praxen-ab-sofort-allgemein-verfuegbar/) | | **Dragon Copilot** | Cloud clinical AI (dictation + ambient) | ✅ Sold, generally available in Germany | 7 Oct 2025 (Germany GA) | [Microsoft newsroom](https://news.microsoft.com/de-de/nach-erfolgreicher-pilotphase-microsoft-dragon-copilot-fuer-deutschlands-kliniken-und-praxen-ab-sofort-allgemein-verfuegbar/) | Each product above gets a one-line definition the first time it matters to you: - **Dragon Professional 16:** the current Windows desktop edition, bought once (perpetual licence) and running recognition fully on the PC. - **Dragon Professional Anywhere:** a cloud subscription with a thin Windows client that sends audio over an encrypted connection to Microsoft Azure. - **Dragon Legal Anywhere:** the legal cloud edition for Germany, hosted in the Deutsche Telekom Open Telecom Cloud rather than Azure. - **Dragon Medical One:** Nuance's cloud dictation for clinicians, in Germany offered as "German Hosted" on Azure in Frankfurt am Main and Berlin. - **Dragon Copilot:** Microsoft's clinical assistant that combines Dragon Medical One dictation with ambient listening from DAX Copilot. ![Archive shelf with old unbranded software boxes, symbolising discontinued Dragon editions](https://api.sprecho.ai/media/2026/10/1791196857-1aa2c519.jpg) ## How Nuance retires cloud versions (and why dates keep moving) **Every Dragon cloud release has a fixed 24-month life: 12 months supported, 6 months limited support, then end of life.** According to the Dragon Legal Anywhere release notes, an end-of-life version "will no longer function due to advances in the Nuance cloud infrastructure", so staying on an old client is not an option. - **Dragon Professional Anywhere 2024.2:** end of life on 31 August 2026 ([release notes 2026.1](https://supportcontent.nuance.com/kbfiles/DPA_RELEASE_NOTES_2026.1-en.htm)). - **Dragon Professional Anywhere 2024.3:** end of life on 30 September 2026. - **Dragon Professional Anywhere 2025.1:** end of life on 13 February 2027. - **Dragon Legal Anywhere 2024.1:** end of life on 20 February 2026 ([release notes 2024.2](https://supportcontent.nuance.com/kbfiles/DLA_RELEASE_NOTES_2024.2.htm)). - **Dragon Medical One LTSR:** an integrator's migration report states that long-term service builds 2021.4.5 and earlier stop working after 31 December 2026; Nuance has not published that notice openly, so confirm it with your reseller. The product-level end dates are a separate question. Two authorised partners in Germany, Thax Software (notice of 3 September 2026) and Byte Heroes (15 July 2026), report that Dragon Professional Anywhere and Dragon Legal Anywhere stop selling on **31 December 2026** and reach end of life on **31 December 2027**, with renewals not planned under the current information. As of October 2026 we found no matching public advisory from Nuance or Microsoft, so treat these dates as announced, not official. ## Which platforms does Dragon still support? **In October 2026, every Dragon product you can still buy runs on Windows; macOS, Linux and mobile are not covered by any current Dragon dictation product.** The platform matrix shows the gap, with Sprecho as the reference row. | Product | Windows | macOS | Linux | iOS | Android | Web | |---|---|---|---|---|---|---| | Dragon Professional 16 | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ | | Dragon Legal 16 | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ | | Dragon Professional Anywhere | ✅ (client) | ❌ | ❌ | ❌ | ❌ | ✅ | | Dragon Legal Anywhere | ✅ (client) | ❌ | ❌ | ❌ | ❌ | ❌ | | Dragon Medical One | ✅ | ❌ | ❌ | ❌ | ❌ | Tablet browser only | | Dragon for Mac 6 | ❌ | ❌ since 2018 | ❌ | ❌ | ❌ | ❌ | | Dragon Anywhere | ❌ | ❌ | ❌ | ❌ since 1 Jul 2026 | ❌ since 1 Jul 2026 | ❌ | | **Sprecho** | ✅ | ✅ | ✅ | ✅ | ✅ | Via desktop + mobile apps | - **Apple Silicon:** Dragon Professional does not support ARM-based Windows 11, so it cannot run under Parallels on M1 to M5 Macs. - **Intel Macs with Parallels:** the virtualised microphone adds latency, and Dragon dictates only into Windows apps inside the VM, never into native Mail, Pages or Safari. - **Windows 10:** Microsoft ended Windows 10 support on 14 October 2025; Nuance says that if an issue is caused by Windows 10, the user must move to Windows 11 and Dragon v16 ([Nuance advisory 29940](https://nuance.custhelp.com/app/answers/detail/a_id/29940/)). - **Windows 11 24H2:** users on the ProZ forum report that Dragon 15.6 stopped dictating outside DragonPad after the 24H2 update; this is a user report, not a Nuance statement. - **Linux:** never supported by any Dragon edition. ## Why is Dragon being wound down? **Since Microsoft bought Nuance in March 2022, the investment has gone into clinical AI, while consumer, Mac and mobile editions were closed.** Nobody at Microsoft has called it a wind-down, but the release record speaks for itself. - **Desktop pause:** Dragon Professional 16 (February 2023) is the only major desktop release since the acquisition, and no version 17 has been announced in 3.5 years. - **Healthcare focus:** Microsoft merged Dragon Medical One and DAX Copilot into Dragon Copilot and has shipped 13 releases since September 2025. - **Copilot in Germany:** Dragon Copilot became generally available for German clinics and practices on 7 October 2025 ([Microsoft newsroom](https://news.microsoft.com/de-de/nach-erfolgreicher-pilotphase-microsoft-dragon-copilot-fuer-deutschlands-kliniken-und-praxen-ab-sofort-allgemein-verfuegbar/)), but as of version 5.1 (September 2026) the multilingual recording feature set ([Microsoft support](https://support.microsoft.com/en-US/dragon-copilot/physicians/multilingual-conversations)) is documented for the US and Canada, with no published timeline for Germany or Austria. - **Mobile exit:** Dragon Anywhere was the only Nuance mobile dictation product; with its end on 1 July 2026 there is no Nuance app for phones any more. - **Cloud consolidation:** partner notices for the end of Professional Anywhere and Legal Anywhere point the same direction: fewer general-purpose products, more Microsoft clinical platform. ## What the status means if you use Dragon today **Your urgency depends on which edition you run: Mac, mobile and v14 users are already unsupported, cloud subscribers face hard dates, and Dragon Professional 16 users can keep working but should plan.** ### Dragon Professional 16 on Windows - **Status:** sold and supported; your perpetual licence keeps working. - **Strength:** recognition runs 100 % on your PC, which is Dragon's strongest privacy point. - **Risk:** no v17 roadmap, Windows-only, and no matching mobile app since Dragon Anywhere ended. - **Price context:** $699.99 one-time in the US; German resellers list roughly 446 to 854 € netto depending on edition, with the Individual ESD typically around 699 € netto. ### Dragon 14 or 15 - **Dragon 14:** all support ended on 1 September 2023; Nuance warns that unsupported versions "may increase your exposure to security risks or impact your ability to meet compliance obligations". - **Dragon 15:** no longer sold since 27 February 2023; the upgrade to v16 costs about 449 € netto at German resellers. - **Profiles:** v15 profiles must be upgraded to run in v16; for v13 and v14 profiles Nuance recommends exporting the vocabulary first. ### Mac and Dragon Anywhere users - **Mac:** Dragon for Mac has been gone since 22 October 2018, and no current Dragon product dictates into native macOS apps. - **Dragon Anywhere:** the App Store notice reads "As of July 1, 2026, Dragon Anywhere Mobile is no longer available for sale. Purchasing new subscriptions and renewing existing subscriptions is no longer possible." - **Server shutdown:** no date has been published for when the Dragon Anywhere service stops; running subscriptions simply cannot be renewed. ### Professional Anywhere, Legal Anywhere and Medical One subscribers - **Professional and Legal Anywhere:** plan for the partner-announced end of sale on 31 December 2026 and end of life on 31 December 2027, and get your own contract end date in writing from your reseller. - **Legal Anywhere pricing:** roughly 579 to 643 € netto per year depending on reseller (first-year promotions from 298.80 €). - **Medical One:** German reseller quotes run 60 to 70 € per user per month netto on 12- to 36-month terms; moving to Dragon Copilot also means moving administration from Nuance Command Center to Dragon Admin Center. - **Jurisdiction:** Azure Frankfurt and Berlin are certified (C5, ISO 27001), but Microsoft is a US company, so the US CLOUD Act still applies regardless of where the data centre is. ## If you use X, do Y: a decision table **Use this table to turn the status into a next step.** The deadline column repeats the hardest date that applies to your edition. | If you use | Status | Hardest date | Do this | |---|---|---|---| | Dragon 14 | Unsupported | Since 1 Sep 2023 | Export your vocabulary now and move off; do not run it on confidential data | | Dragon 15 | Not sold, supported on Windows | Windows 10 issues require v16 + Windows 11 | Decide between paid v16 upgrade and a cross-platform successor | | Dragon Professional 16 | Sold | Not published | Keep it if Windows-only works; add a mobile and Mac solution | | Dragon for Mac | Discontinued | Since 22 Oct 2018 | Switch to a native macOS dictation app | | Dragon Anywhere | No sales or renewals | Your subscription end date | Pick a mobile app that shares one dictionary with your desktop | | Dragon Professional Anywhere | Sold, partner-announced end | 31 Dec 2026 end of sale (partners) | Ask your reseller for the contract end date; test a replacement in Q4 | | Dragon Legal Anywhere | Sold, partner-announced end | 31 Dec 2026 end of sale (partners) | Check § 203 StGB and BRAO fit of the successor before renewal season | | Dragon Medical One | Sold, merging into Copilot | LTSR 2021.4.5 and older after 31 Dec 2026 | Plan the Copilot migration or a dictation-only alternative | ## Where Dragon is still the stronger product **Dragon still leads in voice-command depth, fully offline desktop recognition and integration with hospital systems.** If these points are decisive for you, keep Dragon where it fits. - **Command grammar:** "Select ", "Correct That", "Cap That", "Insert Before/After", MouseGrid and custom commands go further than any modern AI dictation tool. - **Offline:** Dragon Professional 16 recognises speech entirely on the PC, with no network at all. - **Fachsprache:** three decades of German legal and medical vocabulary are built in. - **Hospital integration:** Dragon Medical One has deep PVS and KIS integrations; Dragon Copilot adds ambient documentation that a dictation app does not replace. ## Sprecho: the modern successor for German professionals **Sprecho is a voice dictation app from Melo Designer GmbH in Lastrup, Germany, that runs on Windows, macOS, Linux, iOS and Android and processes speech only in the EU.** It covers the platforms Dragon has left and keeps the data with a German processor. ![Bright modern office with laptop and smartphone for cross-platform dictation](https://api.sprecho.ai/media/2026/10/1791196857-f454915c.jpg) - **Every platform:** native desktop apps for Windows, macOS and Linux plus mobile apps for iOS and Android, with 3 devices per licence. - **Any app:** push-to-talk hotkey (hold or lock mode) dictates into Word, Outlook, Teams, DATEV, RA-MICRO, browser forms and PVS/KIS text fields. - **EU-only processing:** STRATO servers in Germany, GPU inference by Media Trooper (Germany/Netherlands), no US sub-processor in the dictation path. - **Compliance:** German GmbH as processor, AVV/DPA on Enterprise, TÜV SÜD-certified data protection officer; suited to § 203 StGB professions and GDPR Art. 9 health data. - **Your vocabulary:** a [personal and team dictionary](https://sprecho.ai/help/formatting-and-ai/dictionary) with misspelling rules, plus [snippets](https://sprecho.ai/help/formatting-and-ai/snippets) for boilerplate and [voice commands](https://sprecho.ai/help/formatting-and-ai/voice-commands) for punctuation, § and "scratch that". - **100+ languages:** automatic language detection plus AI formatting and per-app style. - **Pro:** €10.92 per seat per month net, or €9.24 per month billed yearly (€110.82 per year net); for individuals and teams of any size. - **Enterprise:** €21.84 per seat per month net, or €18.47 billed yearly (€221.64 per year net); adds SSO/SAML, MDM deployment, audit-log export, licence management and invoice billing. - **On-premise:** on request for customer-hosted deployments of roughly 50 seats and more. - **Honest limit:** Sprecho has no MouseGrid or select-and-correct grammar; its AI formatting removes most of the need for explicit commands. - **Honest limit:** Sprecho is dictation, not ambient clinical note-taking like Dragon Copilot. - **Try it:** Free Forever plan with 2,000 words per week, or a 14-day Pro trial without a card; see [Sprecho pricing](https://sprecho.ai/pricing). For a feature-by-feature view, read [Sprecho vs Dragon](https://sprecho.ai/compare/sprecho-vs-dragon); for the ranked field, see our [best Dragon alternatives for 2026](https://sprecho.ai/blog/best-dragon-alternatives-2026). ## How to move from Dragon in five steps **A Dragon migration takes an afternoon if you start with the vocabulary.** Your custom words are the real asset; the rest is habit. 1. **Export your words:** in Dragon open DragonBar > Settings > Profile > Open User Profile, then Tools > Vocabulary Center > "Export custom word and phrase list" as TXT or XML ([Nuance help](https://www.nuance.com/products/help/dragon15/dragon-for-pc/enx/dmpe/Content/Vocabulary/exporting_and_importing_a_list_of_custom_words_and_phrases.htm)). 2. **Bring them to Sprecho:** add key terms in the dictionary, or send us your exported list and we convert it. 3. **Rebuild auto-texts as snippets:** signatures, addresses and standard paragraphs become spoken trigger phrases with {date} and {time} variables. 4. **Share with the team:** on Pro or Enterprise, shared dictionaries and snippets keep a practice or firm consistent. 5. **Run both for two weeks:** keep Dragon installed during the trial and compare on your own documents. ### Where to start by profession - **Doctors and clinics:** [Sprecho for healthcare](https://sprecho.ai/solutions/healthcare) covers Arztbriefe, Befunde and PVS fields. - **Lawyers:** [Sprecho for law firms](https://sprecho.ai/solutions/legal) covers Schriftsätze, RA-MICRO and beA browser forms. - **Tax advisors:** [Sprecho for tax consultants](https://sprecho.ai/solutions/tax-consultant) covers DATEV and client correspondence. - **Compliance check:** our [GDPR-compliant dictation software comparison](https://sprecho.ai/blog/best-gdpr-compliant-dictation-software-2026) explains the CLOUD Act and § 203 StGB questions in detail. ## How we keep this page current **We re-check every row against Nuance advisories, release notes, the App Store and Microsoft announcements each quarter.** The next scheduled review is January 2027. - **Official first:** Nuance support advisories, Nuance release notes, the Apple App Store and Microsoft announcements are the primary sources. - **Partner notices labelled:** dates that only resellers have published are marked as partner-announced. - **No guesses:** where no date exists, the table says "not published". ## Conclusion Dragon is not dead, but it is now a Windows desktop product plus Microsoft-hosted cloud services, and the cloud subscriptions for general and legal users are heading for a partner-announced end of sale on 31 December 2026. If you dictate on a Mac, on a phone, or under § 203 StGB and want an EU-only processor, the replacement decision is already due. **Sprecho is the modern, EU-hosted, cross-platform successor to Dragon: start a 14-day free trial and bring your Dragon word list along.** Compare the details on [Sprecho vs Dragon](https://sprecho.ai/compare/sprecho-vs-dragon). ## FAQ ### Is Dragon NaturallySpeaking discontinued? The name is: version 13 (2014) was the last major release called Dragon NaturallySpeaking, and the line continued as Dragon Professional. Dragon Professional 16 for Windows is still sold today, while Dragon Home and Dragon for Mac are discontinued. ### Is Dragon Professional 16 still sold? Yes. Dragon Professional 16 for Windows, released on 28 February 2023, is still sold as a perpetual licence for $699.99 in the US and roughly 446 to 854 € netto at German resellers. It runs fully on-device, which remains its strongest privacy point. ### Will there be a Dragon 17? Nuance has not announced one as of October 2026, 3.5 years after Dragon Professional 16 shipped. Microsoft's visible Dragon investment since the 2022 acquisition has gone into Dragon Copilot for healthcare. ### Is Dragon Anywhere discontinued? Yes. The App Store listing states that as of 1 July 2026 Dragon Anywhere Mobile is no longer available for sale and existing subscriptions cannot be renewed. Nuance has not published a date for switching off the service itself. ### Is Dragon Medical One discontinued? No. Dragon Medical One is still sold and is being merged into Dragon Copilot, which became generally available in Germany on 7 October 2025. Individual releases reach end of life on a rolling basis, and an integrator report states that LTSR builds 2021.4.5 and earlier stop working after 31 December 2026. ### Is Dragon available for Mac? No. Nuance discontinued Dragon Professional Individual for Mac on 22 October 2018, and Dragon Professional 16 does not run on ARM-based Windows 11, so Parallels on Apple Silicon is not an option either. Sprecho offers a native macOS app that dictates into Mail, Pages, Safari and every other app. ### Who owns Dragon now? Microsoft owns Dragon since it completed the acquisition of Nuance on 4 March 2022 for about $19.7 billion. Because Microsoft is a US company, the US CLOUD Act applies to Dragon cloud services even when they are hosted in German Azure data centres. ### What is the best Dragon successor for Germany? For professionals bound by § 203 StGB or handling GDPR Art. 9 health data, Sprecho is the closest modern successor: a German GmbH, EU-only processing, apps for Windows, macOS, Linux, iOS and Android, and Pro from 10.92 € netto per month. Dragon remains stronger for offline-only Windows use and deep voice-command control. --- ## https://sprecho.ai/blog/what-dragon-really-costs-2026-licences-upgrades-medical-legal — What Dragon Really Costs in 2026: Licences, Upgrades, Medical, Legal and VAT Explained > Dragon prices 2026: Professional 16, Legal, Medical One and Anywhere with netto/brutto, upgrades, VAT and a 3-year cost comparison for 1, 5 and 20 seats. *Published: 2026-11-03T18:27:01.096000Z* *Tags: dragon, pricing, comparison, voice-dictation, germany, gdpr, alternatives* **Dragon price 2026** is a harder question than it looks: there is no single price list, most German offers come from resellers, and the same Dragon Professional 16 licence is quoted anywhere between roughly 446 and 854 € netto. This guide collects every current Dragon product with its real price, licence model, term and VAT treatment, shows the hidden costs that never appear on the product page, and works out a transparent 3-year total cost of ownership for 1, 5 and 20 seats, compared with Sprecho. *Last verified: October 2026.* - **Dragon Professional 16 (Windows, perpetual):** $699.99 in the US; German reseller quotes roughly 446 to 854 € netto, with the Individual ESD typically listed around 699 € netto. - **Dragon Medical One (cloud, healthcare):** reseller quotes of 60 to 70 € per user per month netto, depending on a 12, 24 or 36-month term. - **Dragon Legal Anywhere (cloud, legal):** from 298.80 € netto in a discounted first year up to 643.20 € per year netto regular, depending on the reseller. - **Dragon Anywhere (mobile):** no longer sold since 1 July 2026; renewals are impossible. - **Sprecho Pro:** 12.99 € per month brutto, or 10.99 € per month brutto billed yearly (131.88 € per year, 110.82 € netto), on Windows, macOS, Linux, iOS and Android. ## How much does Dragon cost in 2026? The short answer A single Dragon Professional 16 desktop licence costs about **699 € netto** as a one-time payment in Germany, but the real figure for a business is higher once upgrades, maintenance, hardware and a second tool for Mac and mobile are added. Dragon's professional successors are **cloud subscriptions** with minimum terms of 12 to 36 months. - **One price, many quotes:** Nuance (a Microsoft company since 4 March 2022) sells most German business licences through resellers, so the same SKU carries different prices in different shops. - **Netto is the default:** almost every German reseller lists Dragon prices netto, i.e. excluding 19 % VAT. A 699 € netto licence is 831.81 € brutto on the invoice. - **Perpetual is now the exception:** only the Windows desktop editions (Dragon Professional 16 and Dragon Legal 16) are still sold as perpetual licences; everything else is a subscription. - **Prices are not the whole story:** a perpetual licence without upgrades ages; Nuance itself warns that unsupported versions can create security and compliance exposure. ## Every Dragon product and its price (October 2026) The table below lists every Dragon product a German professional can still encounter, with the **licence model**, the **platform** and the **price as quoted by resellers**. All euro amounts are netto unless marked otherwise; your own quote can differ. | Dragon product | Licence model | Platform | Price (reseller quotes) | Status October 2026 | |---|---|---|---|---| | Dragon Professional 16 | Perpetual, one-time | Windows only | $699.99 (US); about 446 to 854 € netto in Germany; Individual ESD around 699 € netto | Sold; current version since 28 February 2023 | | Dragon Professional 16 subscription | 1-year subscription | Windows only | 466.57 € netto per year (396.81 € at 20+ users) at one German reseller | Sold | | Upgrade v15 → Dragon Professional 16 | One-time upgrade | Windows only | About 392 to 449 € netto depending on reseller and seat tier | Sold | | Dragon Legal 16 | Perpetual, volume tiers | Windows only | 1,007.56 € netto per seat (1 to 9 seats) down to 491.19 € (1,001+ seats); upgrade from v15 449 € netto | Sold (Germany and USA only) | | Dragon Legal Anywhere | Cloud subscription | Windows client | 298.80 € netto first year with a promo code; 578.77 to 643.20 € netto per year regular; 69 € netto per month on a 24-month term at another reseller | Sold; end of sale 31 Dec 2026 and end of life 31 Dec 2027 (partner-announced) | | Dragon Professional Anywhere | Cloud subscription | Windows client + web | 59 € netto per month, 24-month minimum; about 494.62 € netto per year at another reseller | Sold; end of sale 31 Dec 2026 and end of life 31 Dec 2027 (partner-announced) | | Dragon Medical One | Cloud subscription | Windows, browser on tablets | 70 € (12 months), 65 € (24 months), 60 € (36 months) netto per user per month; US list $79 to $99 | Sold; merging into Dragon Copilot | | Dragon Anywhere (mobile) | Subscription | iOS, Android | Was $14.99 per month or $149.99 per year | ❌ Not sold since 1 July 2026 | | Dragon Home / Dragon for Mac | Perpetual | Windows / macOS | No longer available | ❌ Discontinued (Home 2023, Mac 2018) | - **Sources:** the German reseller figures come from public price lists checked in October 2026 (software-express.de, the STP marketplace, softguide.de and others); the Dragon Anywhere notice is on the [official App Store listing](https://apps.apple.com/us/app/dragon-anywhere/id1024652126); the Mac discontinuation is documented in [Nuance's own support article](https://nuance.custhelp.com/app/answers/detail/a_id/27843/). - **Reseller-dependent:** treat every euro figure as a quote, not a list price. The spread between the cheapest and most expensive offer for the same licence regularly exceeds 300 €. - **Anywhere editions end in 2027 (partner-announced):** two authorised Nuance partners in Germany, Thax Software (3 September 2026) and Byte Heroes (15 July 2026), announce end of sale on 31 December 2026 and end of life on 31 December 2027 for Dragon Professional Anywhere and Dragon Legal Anywhere. Existing contracts can be renewed one last time in 2026 and then run out in 2027, and there is no direct Nuance successor for the Anywhere editions. There is no public Nuance advisory as of October 2026, which only publishes per-version lifecycles (release notes 2026.1), so confirm in writing before signing: a 24-month minimum term signed now runs past the announced end of life. - **Medical pricing is not public:** Dragon Medical One and Dragon Copilot are sold only through certified resellers in Germany; the 60 to 70 € range is what resellers quote, not a Microsoft price list. ### Dragon Professional 16: what the licence buys **Dragon Professional 16** is Nuance's Windows desktop dictation software, released on 28 February 2023 and still the current version in October 2026. It is the only Dragon product most individual professionals can still buy outright. - **Perpetual licence:** you pay once and may use that version indefinitely, but without updates to a new major version. - **Fully on-device:** recognition runs 100 % locally, which is Dragon's strongest privacy argument and worth acknowledging. - **Six languages:** English, German, French, Spanish, Dutch and Italian. - **Price corridor:** volume tiers at one reseller sit at 784.93 € netto (1 to 9 users) and 706.43 € netto (10 to 50 users); a single-user ESD is listed there at 827.41 € netto, while other shops sell the Individual ESD around 699 € netto. ### Dragon Legal 16 and Dragon Legal Anywhere **Dragon Legal** is the law-firm edition with legal vocabulary, available only for Germany and the USA. It comes as a perpetual Windows licence (Dragon Legal 16) or as a cloud subscription (Dragon Legal Anywhere) hosted in the Deutsche Telekom Open Telecom Cloud. - **Dragon Legal 16 volume pricing:** 1,007.56 € netto per seat for 1 to 9 seats, falling to 491.19 € at 1,001+ seats. - **Dragon Legal Anywhere, regular:** 643.20 € netto per year on the STP marketplace, 578.77 € netto per year at software-express.de, about 619 € netto per year at other resellers. - **Dragon Legal Anywhere, promotions:** the STP marketplace offers 298.80 € netto for the first year with the code DLA-STP26; the regular price applies from year two. - **Monthly variant:** one reseller lists 69 € netto per month with a 24-month minimum term, which adds up to 1,656 € netto over the minimum term. ### Dragon Medical One and Dragon Copilot **Dragon Medical One** is Nuance's cloud dictation for healthcare, offered in a German-hosted variant on Microsoft Azure in Frankfurt and Berlin; Microsoft is merging it with DAX Copilot into **Dragon Copilot**. German pricing exists only as reseller quotes. - **12-month term:** about 70 € netto per user per month (83.30 € brutto). - **24-month term:** about 65 € netto (77.35 € brutto). - **36-month term:** about 60 € netto (71.40 € brutto), i.e. 2,160 € netto per user over the term. - **Dragon Copilot:** generally available in Germany since 7 October 2025; no public German price, so ask your reseller for a written quote. ![Calculator, notepad and coffee on an office desk while comparing Dragon licence and upgrade costs](https://api.sprecho.ai/media/2026/10/1791196825-432490f2.jpg) ## Perpetual vs subscription: which Dragon model are you actually buying? You are buying a **perpetual licence** only with Dragon Professional 16 or Dragon Legal 16 on Windows; every other current Dragon product is a **subscription with a minimum term**. That distinction drives most of the cost difference over three years. - **Perpetual (Professional 16, Legal 16):** high upfront cost, no recurring fee, but no right to the next major version unless you pay for an upgrade or a maintenance contract. - **Annual subscription (Professional 16 subscription, Legal Anywhere):** lower entry price, continuous access, but the fee recurs every year and stops working when you stop paying. - **Multi-year contracts (Medical One, Professional Anywhere):** the monthly price falls with the term length; you commit to 24 or 36 months up front. - **Mobile:** there is no longer any way to pay for Dragon on a phone, because Dragon Anywhere left the market on 1 July 2026. ## Hidden costs: upgrades, maintenance, hardware and the second tool The hidden costs of Dragon are the **upgrade cycle**, optional **maintenance**, **Windows 11 hardware**, **microphones** and a **second dictation tool** for Mac and mobile users. None of them appears on the product page, and together they can exceed the licence price. ### The upgrade treadmill - **Sales of v15 ended on 27 February 2023** and all support for v14 ended on 1 September 2023, according to a [Nuance product advisory](https://nuance.custhelp.com/app/answers/detail/a_id/29838/). - **Nuance's own warning:** continuing to use unsupported versions "may increase your exposure to security risks or impact your ability to meet compliance obligations". - **Upgrade price:** roughly 392 to 449 € netto per seat to move from v15 to v16; earlier versions generally need a new full licence. - **Cloud contracts with an announced end:** Dragon Professional Anywhere and Dragon Legal Anywhere are partner-announced to reach end of life on 31 December 2027 (Thax Software, Byte Heroes); there is no public Nuance advisory as of October 2026, so confirm in writing before signing. A migration within about a year is then a cost you should plan for now. - **No Dragon Professional 17:** as of October 2026 there has been no new major version for 3.5 years. Nobody can tell you today when, or whether, the next paid upgrade will come. ### Maintenance contracts - **What it costs:** one German reseller lists one year of maintenance and support at 117.07 € netto per user (1 to 9 users) and 105.37 € netto (10 to 50 users). - **What it bought in the past:** customers with a current maintenance contract were eligible for a free upgrade from v15 to v16. - **Is it mandatory?** No. It is an optional insurance against the next upgrade fee, and only pays off if a new version actually ships during the contract. ### Hardware, headsets and Windows 11 - **Windows 11 requirement:** Microsoft ended Windows 10 support on 14 October 2025 ([Microsoft end-of-support page](https://www.microsoft.com/en-us/windows/end-of-support)). Nuance states that if an issue is caused by Windows 10, the user must upgrade to Windows 11 and Dragon v16 ([Nuance advisory](https://nuance.custhelp.com/app/answers/detail/a_id/29940/)). - **Microphones:** many Dragon deployments budget for dedicated headsets or a dictation microphone such as Nuance PowerMic; prices vary widely, so we leave them out of the arithmetic. - **Older PCs:** a forced move to Windows 11 can mean new hardware for workstations that do not meet its requirements. ### A second tool for Mac and mobile users - **Mac:** Dragon for Mac was discontinued on 22 October 2018, and Dragon Professional does not support ARM-based Windows 11, so it does not run in Parallels on Apple Silicon. - **Mobile:** Dragon Anywhere can no longer be bought or renewed since 1 July 2026. - **Linux:** never supported. - **Budget effect:** a mixed Windows and Mac office pays for Dragon on the PCs plus a separate product on every Mac and phone, with two vocabularies to maintain. ### Reseller services - **Setup and training:** resellers may charge for installation, profile setup, training or admin onboarding. We do not quote numbers because they differ per reseller; ask for them in writing. - **Medical and legal integrations:** connecting Dragon Medical One to a practice or hospital system is typically a reseller project with its own cost. ## Netto, brutto and VAT: what your invoice will actually say German Dragon prices are almost always quoted **netto**, so a German business receives an invoice with **19 % VAT added** and reclaims it as input tax, while a private buyer pays the brutto amount. Sprecho shows **brutto prices** on its pricing page and breaks out the net amount at checkout. | Purchase route | Who invoices | VAT for a German buyer | Reverse charge? | Pay by invoice? | |---|---|---|---|---| | Dragon via a German reseller (perpetual or subscription) | The reseller | 19 % added on top of the netto quote | No, domestic supply | Usually yes, check terms | | Any software supplier based in another EU country (general rule) | The foreign supplier | 0 % on the invoice; the German business accounts for VAT itself | Yes, for businesses with a valid VAT ID | Depends on supplier | | Dragon Medical One / Dragon Legal Anywhere contract | The certified reseller | 19 % added to the monthly or yearly netto fee | No, domestic supply | Usually yes, fixed term | | Sprecho Pro (Germany) | Melo Designer GmbH | 19 % included in the displayed brutto price | No, domestic supply | No, card, SEPA or PayPal | | Sprecho for an EU business outside Germany | Melo Designer GmbH | 0 % with a verified VAT ID | Yes (VIES-verified) | Enterprise only | | Sprecho Enterprise | Melo Designer GmbH | 19 % shown separately at checkout | Only outside Germany | ✅ Invoice and bank transfer, 14 days | - **Compare like with like:** 699 € netto (Dragon) and 12.99 € brutto (Sprecho Pro monthly) are not the same kind of number. Convert both to netto before comparing: Sprecho Pro is 10.92 € netto per month, or 9.24 € netto on yearly billing. - **VAT ID does not remove German VAT:** a German business with a German VAT ID still pays 19 % on a domestic supply; reverse charge only applies across borders. Sprecho explains this in its [VAT and business billing guide](https://sprecho.ai/help/plans-and-billing/vat-and-business-billing). - **Ask for the full quote:** for any reseller offer, request the netto amount, the VAT line, the minimum term and any setup fees in one document. ## 3-year total cost of ownership: 1, 5 and 20 seats Over three years, **Sprecho Pro is the cheapest option for every team size** in this comparison, Dragon Professional 16 is mid-range if you never upgrade, and Dragon Medical One is the most expensive because it is a 36-month cloud contract. Every number below is netto and rests on the labelled assumptions. ### Assumptions (read these first) 1. **Netto throughout:** all amounts exclude 19 % VAT, as a German business that reclaims input tax would see them. 2. **Dragon Professional 16 licence:** 699 € for 1 seat (typical Individual ESD listing), 784.93 € per seat for 5 seats (1 to 9 tier) and 706.43 € per seat for 20 seats (10 to 50 tier), all reseller quotes from October 2026. 3. **Upgrade scenario (assumption, not an announcement):** one paid upgrade at 449 € per seat within three years. No Dragon Professional 17 exists today; the scenario models what happened with v14 and v15, where support ended and an upgrade became the safe option. 4. **Dragon Professional 16 subscription:** 466.57 € per seat per year, or 396.81 € at 20+ users. 5. **Dragon Medical One:** the cheapest observed reseller rate, 60 € per user per month on a 36-month term. 6. **Sprecho:** list prices with yearly billing, Pro 110.82 € and Enterprise 221.64 € per seat per year, no discounts. 7. **Anywhere editions not modelled:** a 36-month line for Dragon Professional Anywhere or Dragon Legal Anywhere would be hypothetical, because both are partner-announced to end on 31 December 2027 (no public Nuance advisory as of October 2026; confirm in writing before signing). The "Dragon Professional 16 subscription" row is the desktop term licence, which the same partners describe as remaining available. 8. **Excluded everywhere:** headsets, hardware, Windows 11 refreshes, staff time, training, reseller services and maintenance contracts. | Option (3 years, netto) | 1 seat | 5 seats | 20 seats | Per seat per month | |---|---|---|---|---| | Dragon Professional 16, licence only | 699.00 € | 3,924.65 € | 14,128.60 € | 19.42 to 21.80 € | | Dragon Professional 16 + one upgrade (assumption) | 1,148.00 € | 6,169.65 € | 23,108.60 € | 31.89 to 34.28 € | | Dragon Professional 16 subscription | 1,399.71 € | 6,998.55 € | 23,808.60 € | 33.07 to 38.88 € | | Dragon Medical One (36-month term) | 2,160.00 € | 10,800.00 € | 43,200.00 € | 60.00 € | | **Sprecho Pro (yearly)** | **332.46 €** | **1,662.30 €** | **6,649.20 €** | **9.24 €** | | Sprecho Enterprise (yearly) | 664.92 € | 3,324.60 € | 13,298.40 € | 18.47 € | ### The arithmetic, line by line - **Dragon Professional 16, licence only:** 1 × 699 € = 699 €; 5 × 784.93 € = 3,924.65 €; 20 × 706.43 € = 14,128.60 €. - **Plus one upgrade:** add 449 € per seat: 699 + 449 = 1,148 €; 3,924.65 + 5 × 449 = 6,169.65 €; 14,128.60 + 20 × 449 = 23,108.60 €. - **Dragon subscription:** 466.57 € × 3 years = 1,399.71 € per seat; × 5 = 6,998.55 €; at 20 seats 396.81 € × 3 × 20 = 23,808.60 €. - **Dragon Medical One:** 60 € × 36 months = 2,160 € per user; × 5 = 10,800 €; × 20 = 43,200 €. - **Sprecho Pro:** 110.82 € × 3 years = 332.46 € per seat; × 5 = 1,662.30 €; × 20 = 6,649.20 €. - **Sprecho Enterprise:** 221.64 € × 3 years = 664.92 € per seat; × 5 = 3,324.60 €; × 20 = 13,298.40 €. - **Per seat per month:** the 3-year total divided by 36 months, e.g. 699 € ÷ 36 = 19.42 €, 332.46 € ÷ 36 = 9.24 €. ### How to read the result - **Dragon licence-only is not "free after year one":** for 1 seat, 699 € is still more than twice the 332.46 € of three years of Sprecho Pro, before any upgrade. - **Brutto view for a private buyer:** 699 € netto is 831.81 € brutto; three years of Sprecho Pro yearly are 3 × 131.88 € = 395.64 € brutto. - **Dragon Medical One buys more than dictation:** deep integrations into hospital and practice systems and, with Dragon Copilot, ambient documentation. If you need those, the comparison with a dictation app is not apples to apples. - **Maintenance would raise Dragon further:** two years of maintenance at 117.07 € add 234.14 € to the 1-seat licence, making 933.14 € netto. - **Platforms are not equal:** each Sprecho seat covers Windows, macOS, Linux, iOS and Android on up to 3 devices; each Dragon seat covers one Windows install. ![Small business owner at a shop counter checking dictation software costs on a laptop](https://api.sprecho.ai/media/2026/10/1791196825-1d5ecd38.jpg) ## What Sprecho costs, and what you get for it Sprecho costs **0 € on the Free plan** (2,000 words per week), **12.99 € per month brutto** on Pro or **10.99 € per month brutto billed yearly**, and 25.99 € or 21.99 € per month brutto on Enterprise. Sprecho is a voice dictation app by Melo Designer GmbH in Lastrup, Germany, processed on servers in Germany and the Netherlands. - **Free Forever:** 2,000 words per week, no card, the full app; see [what the Free plan includes](https://sprecho.ai/help/plans-and-billing/what-the-free-plan-includes). - **Pro:** unlimited dictation, priority support, shared team dictionary and snippets; 10.92 € netto monthly or 9.24 € netto per month billed yearly. - **Enterprise:** admin dashboard, licence management, SSO/SAML, MDM deployment, audit-log export, pay by invoice and an AVV (data processing agreement); 21.84 € or 18.47 € netto per month. - **Trial:** 14 days of Pro, no card required; Enterprise teams get a 14-day test. - **No minimum term:** cancel any time; access runs to the end of the paid period. - **Compare plans:** the full matrix is on the [Sprecho pricing page](https://sprecho.ai/pricing) and in the [plan comparison](https://sprecho.ai/help/plans-and-billing/free-pro-and-enterprise-compared). - **Platforms:** native apps for Windows, macOS and Linux, plus iOS and Android, with 3 devices per licence. - **Any app:** dictate into Word, Outlook, Teams, DATEV, RA-MICRO, browser forms and PVS/KIS text fields with a push-to-talk hotkey. - **100+ languages** with automatic language detection and AI formatting (punctuation, filler removal, per-app style). - **Dictionary and snippets:** add Fachsprache, set capitalisation and misspelling rules; for a long Dragon word list, send us your exported list and we convert it. - **EU-only processing:** STRATO servers in Germany, GPU inference via Media Trooper (DE/NL), no US sub-processor in the dictation path, so the CLOUD Act does not reach your dictations through us. ## Where Dragon is still worth the money Dragon is still worth its price if you need **fully offline recognition**, its **deep voice-command grammar** or the **clinical integrations** of Dragon Medical One. Those are real strengths, and Sprecho does not replicate all of them. - **Offline desktop:** Dragon Professional 16 recognises speech entirely on the PC; Sprecho processes audio on EU servers. - **Command grammar:** "Select", "Correct That", "Insert Before", MouseGrid and custom commands go further than Sprecho's high-frequency command set. - **Clinical depth:** Dragon Medical One integrates deeply with practice and hospital systems; Dragon Copilot adds ambient documentation, which Sprecho does not offer. - **Thirty years of Fachsprache:** Dragon's German legal and medical vocabularies are mature. ## Which option fits your budget? A decision guide Choose by **platform mix, team size and contract appetite**, not by the sticker price alone. The list below maps typical German buyers to the option that usually costs least for what they need. - **Solo professional on one Windows PC, offline is a must:** Dragon Professional 16 perpetual remains defensible; budget for a possible upgrade. - **Solo professional with a Mac or a phone:** Sprecho Pro; Dragon has no current Mac or mobile product. - **Law firm, 5 to 20 seats:** compare Dragon Legal Anywhere quotes with Sprecho Pro or Enterprise; see [Sprecho for law firms](https://sprecho.ai/solutions/legal). - **Medical practice:** if you rely on KIS integration or ambient notes, get a Dragon Medical One or Copilot quote; if you need fast dictation into PVS fields, see [Sprecho for healthcare](https://sprecho.ai/solutions/healthcare). - **Tax advisors and SMB teams:** dictation into DATEV, Outlook and Word on mixed hardware favours Sprecho; see [Sprecho for tax advisors](https://sprecho.ai/solutions/tax-consultant). - **Under § 203 StGB:** whichever product you pick, insist on an AVV and check who controls the processor, not only where the server stands. ## Conclusion Dragon in 2026 costs about 699 € netto for a Windows desktop licence that has not seen a new major version since February 2023, or 60 to 70 € netto per user per month for the medical cloud, with every German price depending on the reseller. Over three years, even the cheapest Dragon route costs about twice as much per seat as Sprecho Pro, and it still leaves Mac and mobile users without a tool. If you need offline recognition or deep clinical integrations, Dragon earns its price; for everyone else, **start the 14-day Sprecho trial without a card** and compare it against your Dragon quote, or read the [full Sprecho vs Dragon comparison](https://sprecho.ai/compare/sprecho-vs-dragon) and our guide to the [best Dragon alternatives in 2026](https://sprecho.ai/blog/best-dragon-alternatives-2026). ## FAQ ### How much does Dragon Professional 16 cost in 2026? Dragon Professional 16 costs $699.99 one-time in the US. In Germany, resellers quote roughly 446 to 854 € netto depending on edition and seat tier, with the Individual ESD typically around 699 € netto (831.81 € brutto). An upgrade from v15 costs about 392 to 449 € netto. ### Is there a Dragon subscription? Yes. Dragon Professional 16 is also sold as a 1-year subscription (466.57 € netto per year at one German reseller), and the cloud products Dragon Professional Anywhere, Dragon Legal Anywhere and Dragon Medical One are subscriptions only, often with 12 to 36-month minimum terms. The mobile subscription Dragon Anywhere can no longer be bought or renewed since 1 July 2026. ### What does Dragon Medical One cost in Germany? There is no public German price list; Dragon Medical One is sold only through certified resellers. Observed reseller quotes are about 70 € netto per user per month on a 12-month term, 65 € on 24 months and 60 € on 36 months, i.e. 71.40 to 83.30 € brutto. Because Microsoft is a US company, the CLOUD Act also applies to its German Azure hosting. ### Is Dragon cheaper than Sprecho over 3 years? No, not on our assumptions. For one seat, Dragon Professional 16 licence-only costs 699 € netto versus 332.46 € netto for three years of Sprecho Pro on yearly billing; with one assumed upgrade Dragon rises to 1,148 €. Dragon Medical One on a 36-month term costs 2,160 € netto per user. ### Are Dragon prices netto or brutto? German Dragon resellers almost always quote netto prices, so 19 % VAT is added on the invoice. A 699 € netto licence is 831.81 € brutto. Sprecho shows brutto prices on its pricing page (12.99 € per month for Pro) and lists the net amount and VAT line separately at checkout. ### Does Sprecho have a free plan? Yes. Sprecho Free Forever includes 2,000 dictated words per week with the full app and needs no credit card. Every new account also starts with a 14-day Pro trial with unlimited words, again without a card. ### Why do Dragon reseller prices differ so much? Nuance sells most German business licences through resellers, who set their own prices, bundle services and run promotions such as first-year discount codes. Volume tiers, ESD versus volume licences and government or education SKUs add further variation. Always request a written netto quote including term and setup fees. ### Do I need a Dragon maintenance contract? No, maintenance is optional; one reseller lists it at 117.07 € netto per user per year for 1 to 9 users. Its main benefit is a free upgrade if a new version ships during the term, as v15 customers with maintenance received v16. As of October 2026 no Dragon Professional 17 exists, so the value depends on future releases nobody has announced. --- ## https://sprecho.ai/blog/migrate-from-dragon-to-sprecho-step-by-step-2026 — How to Migrate from Dragon to Sprecho: Vocabulary, Commands and Shortcuts Step by Step (2026) > Migrate from Dragon to Sprecho in about 30 minutes: export your Dragon vocabulary, rebuild auto-texts as snippets, translate commands and roll out to your team. *Published: 2026-11-03T18:26:57.350000Z* *Tags: dragon, migration, voice-dictation, gdpr, germany, windows, mac* **Migrate from Dragon** to Sprecho without losing the years you invested in your Dragon profile: this guide walks you through exporting your Dragon vocabulary, rebuilding your custom words, auto-texts and commands in Sprecho, setting a push-to-talk hotkey and rolling the result out to a whole firm. For a single user the whole move takes about **30 minutes**; a team of twenty is usually done in an afternoon. *Last verified: October 2026.* **Sprecho** is a voice dictation app from Melo Designer GmbH (Germany) that dictates into any application on Windows, macOS, Linux, iOS and Android, with recognition and AI formatting processed on EU-only infrastructure. **Dragon Professional 16** is Nuance's Windows-only desktop dictation software, released on 28 February 2023 and, as of October 2026, still without a successor version. - **Step 1:** export your Dragon custom words (TXT or XML). - **Step 2:** add them to the Sprecho Dictionary, or send the list to support for conversion. - **Step 3:** rebuild Dragon auto-texts as Sprecho Snippets. - **Step 4:** set your dictation hotkey and choose hold or lock mode. - **Step 5:** learn the command differences (the half-second pause). - **Step 6:** set a writing style per app for Outlook, Word and your practice or case software. - **Step 7:** roll out to your team with a shared dictionary and an AVV. ## Why migrate from Dragon now? Because Dragon's desktop line has stopped moving while the platforms around it have not: there has been no major Dragon version since February 2023, the Mac edition ended in 2018 and the mobile app stopped selling on 1 July 2026. The facts below come from Nuance's own advisories and store listings. - **No Dragon 17:** Dragon Professional 16 shipped on 28 February 2023. Three and a half years later there is still no new major version. - **Old versions are unsupported:** all support for Dragon v14 ended on 1 September 2023, and Nuance warns that unsupported versions "may increase your exposure to security risks or impact your ability to meet compliance obligations" ([Nuance product advisory](https://nuance.custhelp.com/app/answers/detail/a_id/29838/)). - **Windows 10 is over:** Microsoft ended Windows 10 support on 14 October 2025. Nuance says v15 and v16 remain supported there, but if an issue is caused by Windows 10 the user must move to Windows 11 and Dragon v16. - **Mac is gone since 2018:** Dragon Professional Individual for Mac 6 was discontinued on 22 October 2018, and Dragon Professional does not support ARM-based Windows 11, so it cannot run under Parallels on Apple Silicon. - **Mobile is gone since July 2026:** the [App Store listing](https://apps.apple.com/us/app/dragon-anywhere/id1024652126) for Dragon Anywhere states that as of 1 July 2026 it is no longer available for sale and renewals are no longer possible. - **Dragon's successors are cloud subscriptions:** Dragon Professional Anywhere and Dragon Medical One run on Microsoft Azure, a US legal entity subject to the CLOUD Act regardless of data-centre location. To be fair, Dragon still does some things better, and you should know them before you switch: - **Fully offline recognition:** Dragon Professional 16 recognises speech 100 % on-device. Sprecho processes audio on its servers in the EU. - **The richest command grammar:** Select, Correct That, Cap That, Insert Before/After and MouseGrid have no Sprecho equivalent. - **A perpetual licence:** you pay once (about $699.99 in the US, roughly 699 € netto for the Individual ESD at German resellers) instead of a subscription. If those three points outweigh everything else for you, stay on Dragon for now. If you need Mac, mobile, team features or ongoing updates, read on. For the full head-to-head, see our [Sprecho vs Dragon comparison](https://sprecho.ai/compare/sprecho-vs-dragon); for the wider market, our ranking of the [best Dragon alternatives for 2026](https://sprecho.ai/blog/best-dragon-alternatives-2026). ## What do you need before you start? You need three things: access to your Dragon user profile, a Sprecho account (the 14-day Pro trial needs no card) and about half an hour. Nothing has to be uninstalled first; Dragon and Sprecho can run side by side while you test. - **Dragon open with your profile loaded:** the export only works from inside the profile that holds your custom words. - **A list of your auto-texts:** open the Command Browser (DragonBar > Tools > Command Center > Command Browser) and note the custom commands you actually use. - **Sprecho installed:** native apps for Windows, macOS and Linux; on macOS grant Microphone and Accessibility permission. - **Your microphone:** keep the headset you used with Dragon. A wired headset is the cheapest real accuracy upgrade; a quiet room with a modern laptop mic also works. - **A free hotkey:** Sprecho uses Insert on Windows by default. If you use Insert for overtype, plan a different key. ![Hands ticking off a migration checklist in a notebook next to a laptop](https://api.sprecho.ai/media/2026/10/1791196780-e5f366ce.jpg) ## Step 1: How do you export your Dragon vocabulary? You export your Dragon vocabulary from the Vocabulary Center, which writes your custom words and phrases to a TXT or XML file. This is Nuance's documented path ([Nuance help: exporting custom words](https://www.nuance.com/products/help/dragon15/dragon-for-pc/enx/dmpe/Content/Vocabulary/exporting_and_importing_a_list_of_custom_words_and_phrases.htm)): 1. **Open the profile:** in the DragonBar choose **Settings > Profile > Open User Profile** and load your profile. 2. **Open the Vocabulary Center:** choose **Tools > Vocabulary Center**. 3. **Export:** select **Export custom word and phrase list**. 4. **Pick the format:** TXT is the easiest to read and edit; choose XML if you want to keep word properties, as Nuance recommends. 5. **Save it somewhere safe:** this file is the only part of your Dragon investment you will carry forward, so keep a copy. - **Old profiles:** profiles from Dragon v13 or v14 cannot be opened directly in v16; Nuance recommends exporting the vocabulary first. The export above works for that. - **Custom commands:** the Command Browser's Manage mode can import and export commands ([Nuance help: Command Browser](https://nuance.com/products/help/dragon16/dragon-for-pc/enx/dpg-cp/Content/CmdBrowser/about_command_browser.htm)). Sprecho cannot read that file, but it is a useful checklist for Step 3. - **Clean the list:** delete entries you no longer need (former clients, old projects). Most Dragon users find that 50 to 150 words do 90 % of the work. ## Step 2: How do you bring your words into the Sprecho Dictionary? You add your words in Sprecho's Dictionary, one entry at a time, or you send the exported file to Sprecho support and they convert it for you. **There is no self-service bulk import** for Dragon files today, so for long lists the support route is the fastest. ### Add a word 1. Open Sprecho and click **Dictionary** in the sidebar. 2. Click **Add new**. 3. Type the term under **Add a new word** exactly as it should appear, capitalisation included. 4. Click **Add word**. ### Turn recurring errors into misspelling rules Where Dragon kept writing one specific wrong form, use a replacement rule instead of a plain word: 1. In the same dialog, switch on **Correct a misspelling**. 2. Fill in **Misspelling** (what comes out) and **Correct spelling** (what you want). 3. Save. From now on "schmitt" becomes "Schmidt" and "mueller" becomes "Müller" automatically. ### Long lists, pinning and automatic learning - **Long Dragon lists:** send the exported TXT or XML to Sprecho support and they convert it into dictionary entries for you. - **Pin to recognizer:** the recogniser receives up to 200 of the most relevant entries per dictation. Use **Pin to recognizer** on the handful of names you cannot afford to get wrong. - **Learning from corrections:** when you edit a dictation in your history, clear corrections are learned immediately (with Undo) and less certain ones land in **Suggestions** on the Dictionary page. - **Details:** our [dictionary guide](https://sprecho.ai/help/formatting-and-ai/teach-sprecho-your-words-with-the-dictionary) explains every button. ## Step 3: How do you rebuild Dragon auto-texts as snippets? Dragon auto-texts and text-type custom commands become Sprecho **Snippets**: a spoken trigger phrase that expands into your signature, an address or a full paragraph of boilerplate. You say the trigger in the middle of a dictation and the text lands in its place. 1. Click **Snippets** in the sidebar, then **Add new**. 2. Fill in **Trigger phrase**, the words you will say (for example "my closing formula"). 3. Fill in **Expands to**; this is a rich editor for paragraphs, headings, quotes and links. 4. Click **Add snippet**. - **{date}:** inserts the current date, ideal for file notes and letter headers. - **{time}:** inserts the current time, useful for call notes and time records. - **{clipboard}:** inserts whatever is on your clipboard, for example a case number you just copied. - **Trigger choice:** pick two or three words that never occur together by accident; "address" fires mid-sentence, "my postal address" does not. - **Short triggers:** triggers of one to four characters need an exact spoken match; five characters or more tolerate small variations. - **One snippet, every app:** Outlook receives rich text, Slack and WhatsApp receive markdown. More in the [snippets guide](https://sprecho.ai/help/formatting-and-ai/snippets-voice-shortcuts-for-text-you-repeat). ## Step 4: How do you set the push-to-talk hotkey? Sprecho starts dictation with one global key, **Insert on Windows, fn on macOS and Control+Space on Linux**, and you can change it in Settings. Unlike Dragon's always-listening microphone, Sprecho only records while you ask it to. 1. Open **Settings > General**. 2. Under **Keyboard Shortcuts**, click the **Dictation Hotkey** field. 3. Press the key or combination you want; it saves immediately. If it cannot be registered, another program owns it. 4. Optional: add a **secondary shortcut**, for example a foot pedal or a thumb key on an external keyboard. - **Hold-to-talk:** hold the key, speak, release. Best for a sentence, a subject line or a quick reply. - **Lock mode:** tap the key once (under about a third of a second) and recording stays on hands-free; press again to stop and insert. Best for a full letter or a long Befund. - **Cancel:** Escape discards a running recording; nothing is transcribed or inserted. - **Paste again:** Ctrl+Alt+V on Windows and Linux (Control+Cmd+V on macOS) re-inserts your last dictation. - **More:** see [hold-to-talk vs lock mode](https://sprecho.ai/help/dictation/hold-to-talk-vs-lock-mode). ## Step 5: Which Dragon commands work in Sprecho? Punctuation and layout commands carry over almost one to one, "Scratch that" exists, and the selection, correction and mouse commands do not. The table translates the commands Dragon users say most often. | Task | Dragon command | Sprecho (English) | Sprecho (German) | |---|---|---|---| | Full stop | "period" | "period" / "full stop" | "Punkt" | | Comma | "comma" | "comma" | "Komma" | | Colon / semicolon | "colon" / "semicolon" | "colon" / "semicolon" | "Doppelpunkt" / "Semikolon" | | Question / exclamation mark | "question mark" / "exclamation mark" | "question mark" / "exclamation mark" | "Fragezeichen" / "Ausrufezeichen" | | Quotation marks | "open quote" / "close quote" | "open quote" / "close quote" | "Anführungszeichen auf" / "zu" | | Brackets | "open parenthesis" / "close parenthesis" | "open bracket" / "close bracket" | "Klammer auf" / "Klammer zu" | | Section sign § | varies by version | "section sign" | "Paragraphenzeichen" | | Line break | "new line" | "new line" | "neue Zeile" | | Blank line | "new paragraph" | "new paragraph" | "neuer Absatz" / "neuer Abschnitt" | | Undo last phrase | "Scratch that" | ✅ "Scratch that" | ✅ "Streich das" | | Select / correct text | "Select …", "Correct that" | ❌ edit in history, dictionary rules | ❌ im Verlauf bearbeiten | | Capitalise | "Cap that" | ❌ dictionary entry with capitalisation | ❌ Wörterbucheintrag | | Mouse / app control | MouseGrid, "Open Word" | ❌ click, then dictate | ❌ klicken, dann diktieren | - **The half-second rule:** finish your sentence, pause for about half a second, then say the command. Without the pause, "new line" is written as words, because Sprecho decides from context whether it is a command. - **Exact wording:** "new paragraph" is a command, "paragraph" alone is an ordinary noun and stays in the text. - **Stop dictating commas:** Sprecho punctuates automatically from the meaning of the sentence. Only say a mark when you need it exactly there, such as in a statute citation. - **Language matters:** a German command inside an English-configured dictation is treated as text; check your dictation language. - **Reference:** the [full voice command list](https://sprecho.ai/help/formatting-and-ai/dictating-voice-commands-reliably) also covers slash, hyphen and underscore. ## Step 6: How do you set a writing style for Outlook, Word and your PVS? You set it once on the **Models** page: Sprecho recognises which app you are dictating into and applies the formatting model you chose for that category. Dragon applies one set of formatting options to every application. 1. Click **Models** in the sidebar. 2. Choose a tab: **Personal messages**, **Work messages**, **Email**, **Code** or **Other**. 3. Pick **Standard** (keeps your wording, cleans punctuation and layout) or **Thinking** (rewrites for grammar and readability without changing meaning). 4. Repeat per tab. Changes apply to your next dictation. - **Outlook and Gmail (Email):** Thinking, the shipped default, gives fuller prose for client letters. - **Word, practice and case software (Other):** most word processors, PVS/KIS fields, DATEV and RA-MICRO fall into Other. Standard is the faithful choice for Befunde and pleadings; switch to Thinking if you write long free text there. - **Teams and Slack (Work messages):** Standard keeps your short replies sounding like you. - **Per device:** the setting is stored on each device, so a work laptop and a private Mac can differ. See [per-app style settings](https://sprecho.ai/help/formatting-and-ai/set-a-writing-style-per-app). ## Step 7: How do you roll Sprecho out to a whole firm? You roll out with a team: one shared dictionary and shared snippets for everyone, an admin console for seats, and an AVV that names a German GmbH as processor. Teams start on Pro; Enterprise adds the administration features IT departments ask for. - **Shared dictionary:** switch on **Share with team** when adding an entry; it appears under **Shared with team** for every member. - **Who edits what:** owners and admins can edit any shared entry; members can add entries and edit their own. - **Shared snippets:** one wording for the standard Mandantenanschreiben or discharge paragraph across the firm. - **Enterprise:** admin dashboard, licence management, SSO/SAML, MDM deployment, audit-log export and payment by invoice. - **AVV / DPA:** a data processing agreement under Art. 28 GDPR, available in German and English, with Melo Designer GmbH as processor. - **Devices:** each licence covers 3 devices, so a lawyer can use the office PC, a MacBook and a phone. 1. **Collect:** ask three or four heavy Dragon users for their exported lists; the overlap is your first shared dictionary. 2. **Convert:** send the merged list to support for conversion, then mark firm-wide terms as shared. 3. **Pilot:** run a 14-day test with one department. 4. **Deploy:** roll out via MDM (Enterprise) or the standard installers. 5. **Review:** after a month, check usage counts and remove stale entries. Details in [shared team dictionary and snippets](https://sprecho.ai/help/teams-and-enterprise/shared-team-dictionary-and-snippets). ![Small team at a table with laptops during a friendly onboarding session](https://api.sprecho.ai/media/2026/10/1791196781-140c4491.jpg) ## What transfers from Dragon, and what does not? Your words, auto-texts and punctuation habits transfer; your acoustic profile, macros and mouse commands do not. For most professionals the parts that stay behind are the parts that Sprecho no longer needs. | Dragon element | Transfers? | How in Sprecho | Why it is fine | |---|---|---|---| | Custom words and phrases | ✅ | Dictionary entries, or support converts your export | Same vocabulary, also shared with your team | | Recurring misrecognitions | ✅ | Misspelling rules | Fixed automatically on every dictation | | Auto-texts | ✅ | Snippets with {date}, {time}, {clipboard} | Formatting adapts to each app | | Punctuation commands | ✅ | Same phrases, pause before | Mostly unnecessary thanks to auto punctuation | | Scratch That | ✅ | "Scratch that" / "Streich das" | Removes the phrase spoken just before | | Acoustic profile / voice training | ❌ | Not needed | The recogniser works for new speakers without training | | Select / Correct / Cap That | ❌ | Edit in history, dictionary learns | AI formatting removes most manual correction | | Advanced scripting macros | ❌ | No equivalent | Keep Dragon for those few workflows if you rely on them | | MouseGrid / app control | ❌ | No equivalent | Sprecho dictates; you click and navigate as usual | - **No training session:** Dragon builds a per-user acoustic profile and offers training through its Accuracy Center. Sprecho's recogniser is not trained per user, so a new colleague dictates productively on day one. - **No profile to corrupt:** Dragon profiles must be upgraded or rebuilt between major versions. Sprecho keeps your dictionary and snippets in your account across devices. - **100+ languages:** Dragon supports six languages; Sprecho detects the language automatically, which helps with English terms in German letters. - **Honest limit:** if your daily work depends on voice-driven macros or hands-free mouse control, Sprecho is not a drop-in replacement for that part. ## Migration checklist: who does what, and how long does it take? A single user finishes in about 30 minutes; a team adds roughly one hour for the shared list and rollout. Print the checklist and tick it off. | Step | Time | Who | Done | |---|---|---|---| | 1. Export Dragon vocabulary (TXT/XML) | 5 min | Each user | ☐ | | 2. Add top words, misspelling rules, send long list to support | 10 min | Each user | ☐ | | 3. Rebuild 5 to 10 auto-texts as snippets | 5 min | Each user | ☐ | | 4. Set dictation hotkey, try hold and lock mode | 3 min | Each user | ☐ | | 5. Practise the pause before commands | 3 min | Each user | ☐ | | 6. Set per-app style (Email, Other) | 4 min | Each user | ☐ | | 7a. Merge and share firm dictionary and snippets | 30 min | Team admin | ☐ | | 7b. Sign AVV, set up SSO/MDM (Enterprise) | 30 min | IT / DPO | ☐ | - **Run in parallel:** keep Dragon installed for two weeks while you compare results on real letters. - **Measure:** count how many corrections you make per page in week one and week two. - **Decide:** retire Dragon seats only when the team is comfortable. ## What does Sprecho cost after the switch? Sprecho Pro costs **12.99 € per month, or 10.99 € per month billed yearly** (prices brutto incl. 19 % VAT; 10.92 € and 9.24 € netto). Enterprise costs 25.99 € or 21.99 € per month brutto. All plans are listed on [Sprecho pricing](https://sprecho.ai/pricing). - **Free Forever:** 2,000 words per week, no card, enough to test the migration. - **14-day Pro trial:** no card required. - **Compared with Dragon:** Dragon Professional 16 is a one-time purchase (around 699 € netto at German resellers; the upgrade from v15 costs 449 € netto). Dragon Professional Anywhere is quoted by resellers at around 59 € per month netto on a 24-month term. - **On-premise:** available on request for larger deployments. ## Who benefits most from migrating? Firms with mixed Windows and Mac fleets, mobile staff and confidentiality duties benefit most, because Dragon no longer covers Mac or mobile and its cloud successors run on US-owned infrastructure. ### Law firms - **§ 203 StGB and BRAO:** a German GmbH as processor avoids handing client secrets to a US-controlled provider. - **Any app:** dictate straight into RA-MICRO, Word and Outlook. See [Sprecho for law firms](https://sprecho.ai/solutions/legal). ### Medical practices and clinics - **GDPR Art. 9:** health data dictated into a PVS or KIS is special-category data; EU-only processing keeps it out of US jurisdiction. - **Snippets for Befunde:** standard findings paragraphs by voice. See [Sprecho for healthcare](https://sprecho.ai/solutions/healthcare). ### Tax advisors and public administration - **Tax advisors:** dictate into DATEV text fields and client letters; also bound by § 203 StGB. See [Sprecho for tax advisors](https://sprecho.ai/solutions/tax-consultant). - **Public administration:** Linux support and EU-only processing for authorities. See [Sprecho for public administration](https://sprecho.ai/solutions/government). ## Conclusion Migrating from Dragon is less about software and more about carrying over what you built: your words, your auto-texts and your habits. Export the vocabulary, rebuild the snippets, learn the half-second pause and you will be dictating productively in Sprecho within half an hour, on Windows, Mac, Linux and your phone, with processing kept in the EU. If you depend on Dragon's macros or fully offline recognition, keep a Dragon seat for those cases. For everyone else, **start your free 14-day Sprecho Pro trial today and bring your Dragon word list with you**; compare the details first in our [Sprecho vs Dragon comparison](https://sprecho.ai/compare/sprecho-vs-dragon). ## FAQ ### Can I import my Dragon user profile into Sprecho? No, a Dragon user profile contains an acoustic model that only Dragon can read, and Sprecho does not need one. What you can carry over is your vocabulary: export your custom words from Dragon's Vocabulary Center and add them to the Sprecho Dictionary, or send the list to Sprecho support for conversion. ### How do I export my Dragon vocabulary? Open your profile via DragonBar > Settings > Profile > Open User Profile, then choose Tools > Vocabulary Center > Export custom word and phrase list. Dragon saves the list as TXT or XML. For v13 or v14 profiles, Nuance recommends exactly this export before moving to a new version. ### Is there a bulk import for Dragon word lists in Sprecho? Not as a self-service feature today. You can add entries one by one in the Dictionary, which is quick for the 50 to 150 words most people really use. For long lists, send your exported TXT or XML file to Sprecho support and they convert it into dictionary entries for you. ### Does Sprecho learn my voice like Dragon's profile training? Sprecho does not need acoustic voice training, so new users can dictate productively immediately. It learns your vocabulary instead: when you correct a dictation in your history, clear corrections are learned straight away and less certain ones appear as Suggestions on the Dictionary page. ### Does Sprecho support Scratch That? Yes. Say "Scratch that" in English or "Streich das" in German to remove the phrase you spoke immediately before. Dragon's Select, Correct That, Cap That and MouseGrid commands have no equivalent; Sprecho's AI formatting and dictionary rules remove most of the need for them. ### Do I need a headset to dictate with Sprecho? No. A modern laptop microphone in a quiet room works well, and you can keep the headset you used with Dragon. A wired headset is the cheapest real accuracy upgrade in noisy rooms, while Bluetooth headsets are the weakest option because they drop to a lower-bandwidth mode when the microphone is active. ### Can my whole law firm or practice migrate together? Yes. With a team, dictionary entries and snippets can be shared with everyone, owners and admins can edit any shared entry, and Enterprise adds SSO/SAML, MDM deployment, audit-log export and payment by invoice. Each licence covers 3 devices across Windows, macOS, Linux, iOS and Android. ### Is my data safe during and after the migration? Sprecho is operated by Melo Designer GmbH in Germany, servers run at STRATO in Germany and GPU inference at Media Trooper in Germany and the Netherlands, with no US sub-processor in the dictation path. That keeps dictations outside the US CLOUD Act, which matters for § 203 StGB professionals and health data under GDPR Art. 9, and an AVV under Art. 28 GDPR is available. --- ## https://sprecho.ai/blog/windows-10-end-of-support-is-dragon-still-safe-2026 — Windows 10 Is Out of Support: Is Your Dragon Still Safe to Use? (2026) > Windows 10 is out of support. What Nuance says about Dragon 14, 15 and 16 on Windows 10 and 11, the 24H2 reports, GDPR risks and your migration options. *Published: 2026-11-03T18:26:51.075000Z* *Tags: dragon, windows, gdpr, voice-dictation, germany, migration* **Dragon Windows 11 compatibility** became an urgent IT question on 14 October 2025, when Microsoft ended support for Windows 10. If your Kanzlei, Praxis or Behörde still runs Dragon Professional 15 or 16 on Windows 10 machines, you are now working on an unpatched operating system, under a Nuance support policy that pushes every Windows 10 problem towards a paid move to Windows 11 and Dragon v16. This IT guide explains what Nuance officially says, which Dragon version runs where, what the GDPR expects from you, and what your realistic options cost. *Last verified: October 2026.* - **Who this is for:** IT administrators and practice or office managers responsible for dictation workstations in law firms, medical practices, tax offices and public administration. - **What you get:** Nuance's own statements quoted verbatim, a version-by-OS support matrix, a risk table and a decision table with costs. - **Bottom line up front:** Dragon v14 is unsupported, Dragon v15 survives on Windows 10 only on Nuance's conditions, and Dragon v16 is the only Nuance desktop version with a Windows 11 future. ## What changed when Windows 10 support ended? Since 14 October 2025, Windows 10 no longer receives free security fixes, quality updates or technical support from Microsoft. Your Dragon installation keeps starting, but the operating system underneath it is frozen unless you pay for Extended Security Updates. - **End of support date:** Microsoft stopped technical support, feature updates and quality updates (including security fixes) for Windows 10 on 14 October 2025, according to its [Extended Security Updates documentation](https://learn.microsoft.com/en-us/windows/whats-new/extended-security-updates). - **Commercial ESU:** Organisations can buy Extended Security Updates through volume licensing at **$61 per device for year one**; the price doubles every year, for a maximum of three years after end of support. - **Consumer ESU:** For private and Windows 10 Home users, Microsoft now states that you can enrol "any time until the program ends on October 12, 2027" on its [consumer ESU page](https://www.microsoft.com/en-us/windows/extended-security-updates). - **Prerequisite:** ESU only works on Windows 10 version 22H2, so older builds must be updated first. - **What ESU does not do:** It patches Windows only. It does not include technical support, and it does not change anything about Dragon's own support status. That last point is the one most IT teams miss. Buying ESU keeps Windows 10 patched, but it does not make Nuance treat Windows 10 as a first-class platform again. ## What does Nuance officially say about Dragon on Windows 10? Nuance says Dragon v15 and v16 remain supported on Windows 10, but any problem caused by Windows 10 must be solved by moving to Windows 11 and Dragon v16. Here is the exact wording of the support notice "Dragon and Windows 10 end of support" (answer 29940, updated 13 October 2025): > "Dragon v15 and v16 products will continue to be supported in Windows 10 environments. However, if it is determined an issue is caused by the Windows 10 Operating System, the User must upgrade to Windows 11 and Dragon v16 software." ([Nuance support notice 29940](https://nuance.custhelp.com/app/answers/detail/a_id/29940/)) - **Supported, with a condition:** Dragon v15 and v16 are not cut off on Windows 10, so nothing stops working on a fixed date. - **The condition is open-ended:** Nuance decides whether an issue is "caused by the Windows 10 Operating System", and the remedy is always the same: Windows 11 plus Dragon v16. - **v15 owners pay twice:** A v15 user hit by that clause needs both a new or upgraded PC and a Dragon v16 licence. The older product advisory from February 2023 is even more direct about versions that have left support. In answer 29838, Nuance announced that v15 sales ended on 27 February 2023, that "Support for Dragon v14 (all editions and languages) will end on September 1, 2023", and that there will be no more "updates to these products, including security patches and fixes". It then warns: > "If you elect to continue using your Dragon v14 or v15 solution after the end of sale or support date, doing so may increase your exposure to security risks or impact your ability to meet compliance obligations." ([Nuance product advisory 29838](https://nuance.custhelp.com/app/answers/detail/a_id/29838/)) - **Dragon v14:** No support of any kind since 1 September 2023, in any language or edition. - **Dragon v15:** Out of sale since 27 February 2023; general technical support continued, but Nuance's own advisory names the security and compliance exposure. - **Dragon v16:** Released on 28 February 2023 and described by Nuance as "optimized for Windows 11 and compatible with Windows 10". As of October 2026 there is still no Dragon Professional 17. - **Upgrade path:** Customers with a current maintenance contract were eligible for a free v16 upgrade; everyone else pays, at German resellers typically **449 € netto** for the v15 to v16 upgrade. ## Which Dragon version runs on which operating system? Only Dragon Professional 16 has an official Windows 11 path, and no current Dragon desktop version runs on ARM-based Windows, macOS or Linux. The matrix below combines Nuance's notices and the Dragon 16.1 [system requirements](https://www.nuance.com/products/help/dragon161/dragon-for-pc/enx/dpg-cp/Content/ReleaseNotes/DG_relnotes_16.htm), which list Windows 11 (64-bit) and Windows 10 and only Intel or AMD processors. | Version | Windows 10 | Windows 11 (x64) | Windows 11 on ARM | macOS | Linux | |---|---|---|---|---|---| | Dragon v14 | ❌ Support ended 1 Sep 2023 | ❌ Unsupported | ❌ | ❌ | ❌ | | Dragon v15 | ✅ Supported, but Win10-caused issues require Win11 + v16 | Not Nuance's Windows 11 path; users report breakage after 24H2 | ❌ | ❌ | ❌ | | Dragon Professional 16 | ✅ Supported (32 and 64-bit) | ✅ Supported (64-bit) | ❌ Only Intel/AMD listed | ❌ Mac edition discontinued 22 Oct 2018 | ❌ Never supported | | Sprecho | ✅ Windows 10 or later, 64-bit | ✅ Same installer | x64 build: test in the 14-day trial | ✅ Apple Silicon and Intel | ✅ .deb for Debian/Ubuntu | - **Mac is not a workaround:** Nuance stopped sales and support of Dragon Professional Individual for Mac 6 on 22 October 2018 ([Nuance answer 27843](https://nuance.custhelp.com/app/answers/detail/a_id/27843/)), and Dragon cannot run in Parallels on Apple Silicon because that requires ARM-based Windows 11. - **Copilot+ PCs are a trap:** Many new business laptops ship with ARM processors. Dragon's requirements name Intel or AMD only, so check the CPU before you order replacement hardware for dictation users. - **Windows Server:** Dragon 16.1 also lists Windows Server 2016, 2019 and 2022, which matters for Terminal Server setups. ## What do users report about Dragon 15 after the Windows 11 24H2 update? Users report that Dragon Professional 15.6 stopped dictating into anything except DragonPad after their PCs received the Windows 11 24H2 update. These are user reports, not an official Nuance statement, but they match the pattern Nuance's own notices describe: v15 is not the version Nuance positions for Windows 11. - **The reported symptom:** On the ProZ.com translators' forum, users describe Dragon 15.6 that still works in DragonPad but no longer writes into other applications, including browsers, after the 24H2 update. - **An earlier precedent:** In a 2022 thread on [Microsoft Q&A](https://learn.microsoft.com/en-us/answers/questions/73ac9ebb-781b-40b0-9156-4761bafa0e26/solved-dragon-naturallyspeaking-156-stopped), a user describes Dragon 15.61 freezing at "initializing the user profile" after a Windows 11 22H2 update; the community fix was a full removal, clean boot and step-by-step reinstall. - **What it means for IT:** Every Windows feature update becomes a regression test for dictation. With v15 you cannot open a ticket expecting a fix, because Nuance's stated remedy is v16. - **Practical rule:** Hold feature updates for dictation machines in a pilot ring until Dragon has been tested on them, or move those users to a tool that is tested against current Windows builds. ![Laptop showing a system update progress bar during a Windows migration](https://api.sprecho.ai/media/2026/10/1791196824-6173771f.jpg) ## Is running unsupported Dragon a GDPR problem? Yes, it can be. Article 32 GDPR requires technical and organisational measures (TOMs) appropriate to the risk, "taking into account the state of the art", and dictation in law firms and medical practices processes exactly the kind of data where that bar is highest. - **Art. 32 GDPR:** Running an operating system without security updates, or software whose vendor no longer patches it, is hard to defend as state of the art in your TOMs. - **Art. 9 GDPR:** Medical dictation contains health data, a special category with stricter requirements; voice recordings can also count as biometric data when they are processed to identify a person. - **§ 203 StGB:** Doctors, lawyers and tax advisors are bound by professional secrecy under criminal law, so a breach caused by an unpatched dictation PC is not just a data protection issue. - **Nuance's own words:** The vendor itself warns that unsupported versions may "impact your ability to meet compliance obligations". An auditor or data protection officer will read that sentence too. - **Accountability:** Under Art. 5(2) GDPR you must be able to demonstrate compliance, so document which dictation version runs on which OS and why. | Scenario | Risk | What Nuance says | What to do | |---|---|---|---| | Dragon v14 on any Windows | No patches or support since 1 Sep 2023 | Continued use "may increase your exposure to security risks" | Retire now; do not process client or patient data with it | | Dragon v15 on Windows 10 without ESU | Operating system unpatched since 14 Oct 2025 | Supported, but Win10-caused issues require Win11 + v16 | Buy ESU as a bridge or migrate this quarter | | Dragon v15 on Windows 11 24H2 | Users report dictation only works in DragonPad | v16 is the version "optimized for Windows 11" | Pilot-test before rollout; upgrade or replace | | Dragon v16 on Windows 10 + ESU | Patched OS, conditional vendor support | Same Win10 clause applies | Treat as a bridge; schedule the Windows 11 move | | Dragon on ARM laptops or Macs | Does not run (Intel/AMD only; no Mac edition) | Mac edition discontinued 22 Oct 2018 | Keep x64 Windows hardware or switch to a cross-platform tool | ## What are your options, and what do they cost? You have four realistic paths: stay on Windows 10 with ESU, upgrade to Windows 11 and Dragon Professional 16, move to Nuance's cloud subscription, or replace Dragon with a continuously updated tool such as Sprecho. The table compares them on cost, future-proofing, platform reach and updates. | Option | Cost | Future-proof? | Mac / Linux? | Updates | |---|---|---|---|---| | Stay on Windows 10 + commercial ESU | $61/device in year 1, doubling yearly (max. 3 years) | ❌ Bridge only | ❌ | Windows security only; Dragon support conditional | | Windows 11 + Dragon Professional 16 | 449 € netto upgrade from v15; around 699 € netto new (reseller) | Uncertain: no major version since Feb 2023 | ❌ | Service releases (16.1), no v17 | | Dragon Professional Anywhere (cloud) | e.g. 59 € netto per month, 24-month term (reseller quote) | ❌ Partner-announced: end of sale 31 Dec 2026, end of life 31 Dec 2027 (no public Nuance advisory as of Oct 2026; confirm in writing) | ❌ Windows client | Cloud on Microsoft Azure (US parent) | | Sprecho Pro / Enterprise | 10.92 € / 21.84 € net per seat and month (9.24 € / 18.47 € billed yearly) | ✅ Continuously updated | ✅ macOS, Linux, iOS, Android | In-app updates; centrally controlled on Enterprise | ### Option 1: Stay on Windows 10 with ESU - **Good for:** buying time while hardware is replaced, especially if some PCs cannot run Windows 11. - **Watch out:** ESU is cumulative, so starting in year two means paying year one too, and Nuance's Windows 10 clause still applies to every Dragon ticket. ### Option 2: Upgrade to Windows 11 and Dragon Professional 16 - **Good for:** teams that rely on Dragon's deep command grammar or need fully offline recognition. - **Watch out:** you invest in a perpetual licence for a product without a new major version for more than three and a half years, on Windows x64 hardware only. - **Cost check:** the 449 € netto v15 to v16 upgrade alone roughly equals four years of Sprecho Pro billed yearly (110.82 € net per year). ### Option 3: Move to Dragon Professional Anywhere - **Good for:** organisations that want to stay with Nuance and accept a subscription. - **Watch out:** it runs on Microsoft Azure; Microsoft is a US company, so the US CLOUD Act (2018) applies regardless of the data centre location. Reseller prices vary and minimum terms are common. - **Announced end date:** Thax Software (article of 3 September 2026) and Byte Heroes (15 July 2026), both authorised Nuance partners in Germany, announce end of sale on 31 December 2026 and end of life on 31 December 2027 for Dragon Professional Anywhere and Dragon Legal Anywhere, with contracts renewable one last time in 2026 and no direct Nuance successor for the Anywhere editions. This is partner-announced: as of October 2026 Nuance has published no public advisory, only per-version lifecycles (release notes 2026.1: version 2025.1 reaches end of life on 13 February 2027), so confirm the dates in writing before you sign. ### Option 4: Replace Dragon with Sprecho - **Good for:** mixed Windows, Mac and mobile fleets that want a maintained product, EU-only processing and central administration. - **Watch out:** Sprecho needs an internet connection, because recognition runs on servers in Germany rather than on the PC. More detail in our [Sprecho vs Dragon comparison](https://sprecho.ai/compare/sprecho-vs-dragon). ## Where is Dragon still the better choice? Dragon still wins on fully offline recognition and on its very large voice-command grammar, and an honest migration plan should say so. - **Offline desktop recognition:** Dragon Professional 16 recognises speech 100 % on the device, which is its strongest privacy argument. - **Command and control:** "Select", "Correct That", "Cap That", MouseGrid and custom commands go far beyond what Sprecho offers; Sprecho ships the high-frequency punctuation and layout commands and relies on AI formatting for the rest. - **Decades of Fachsprache:** Dragon has three decades of German legal and medical vocabulary behind it. - **Clinical integrations:** Dragon Medical One has deep PVS and KIS integrations that a general dictation tool does not replicate. If none of those four points is decisive for your users, the remaining arguments for Dragon are mostly habit and sunk cost. Our overview of the [best Dragon alternatives in 2026](https://sprecho.ai/blog/best-dragon-alternatives-2026) compares the wider field. ## How does Sprecho fit a Windows 10 and Windows 11 fleet? Sprecho is a dictation app from Melo Designer GmbH in Germany that runs natively on Windows 10 and 11, macOS and Linux, plus iOS and Android, and dictates into any application through a push-to-talk hotkey. For IT, the important part is that it is deployed, updated and governed like any other managed business app. - **One installer for Windows 10 and 11:** Sprecho requires Windows 10 or later, 64-bit, so you can roll it out before your Windows 11 migration is finished. See [Install Sprecho on Windows](https://sprecho.ai/help/getting-started/install-sprecho-on-windows). - **MSI for Intune and Group Policy:** Every release ships a signed .msi next to the .exe; Enterprise admins also get a signed DMG for Jamf and a .deb for Linux in the console. The rollout steps are in [Roll Sprecho out with Intune or Jamf](https://sprecho.ai/help/teams-and-enterprise/roll-sprecho-out-with-intune-or-jamf). - **Update control:** The Enterprise device policy can switch automatic updates off, choose the stable or beta channel or pin a version, and set a forced-update grace period (24 hours by default). - **SSO and SCIM:** SAML 2.0 and OpenID Connect with presets for Microsoft Entra ID, Okta and Google, plus SCIM 2.0 provisioning; leavers lose access and free their licence automatically. Setup guide: [Set up single sign-on](https://sprecho.ai/help/teams-and-enterprise/set-up-single-sign-on-sso-and-saml). - **Terminal Server:** Multi-session Windows environments get their own agent MSI on the same Downloads page. - **Any app:** Word, Outlook, Teams, DATEV, RA-MICRO, browser forms and PVS or KIS text fields, with 100+ languages and automatic language detection. - **Admins see counts, not content:** The Enterprise console shows usage figures and app names, never the text or audio of anyone's dictation; that is enforced on the server. - **EU processor:** Melo Designer GmbH is a German GmbH; servers are operated by STRATO GmbH in Germany, GPU inference by Media Trooper in Germany and the Netherlands, and there is no US sub-processor in the dictation path. - **Contracts:** An AVV (data processing agreement under Art. 28 GDPR) is available on Enterprise, and the data protection officer is TÜV SÜD-certified. - **Pricing:** Pro costs 12.99 € per month or 10.99 € per month billed yearly (gross, incl. 19 % VAT); Enterprise costs 25.99 € or 21.99 € per month (gross). Details on the [Sprecho pricing page](https://sprecho.ai/pricing). - **Try first:** A 14-day Pro trial without a credit card, and a 14-day Enterprise test for organisations. ![Law firm office with several Windows workstations that need a supported dictation setup](https://api.sprecho.ai/media/2026/10/1791196824-09cddf8b.jpg) ## How do you migrate dictation workstations in 30 days? Inventory first, pilot second, and switch off unsupported Dragon installations last. This sequence works whether you end up on Dragon 16 or on Sprecho. 1. **Inventory (week 1):** List every PC with Dragon, its version (v14, v15, v16), Windows build and CPU architecture. Flag v14 installs and ARM devices immediately. 2. **Export vocabularies:** In Dragon, open DragonBar > Settings > Profile > Open User Profile, then Tools > Vocabulary Center > "Export custom word and phrase list" (TXT or XML), per user. 3. **Decide per group:** Users who need offline recognition or MouseGrid stay on Dragon 16; everyone else is a candidate for Sprecho. 4. **Pilot (week 2):** Install Sprecho via MSI on five to ten machines, connect SSO first, and send the exported word lists to Sprecho support, who convert them into dictionary entries. 5. **Update the TOMs (week 3):** Record the new processor, sign the AVV and document the retention setting you chose in the Enterprise console. 6. **Roll out and retire (week 4):** Deploy to everyone else, then uninstall unsupported Dragon versions so that nobody keeps dictating client data into them. ## Who should act first? Anyone processing data covered by § 203 StGB or Art. 9 GDPR on Windows 10 or with Dragon v14 or v15 should act first, because those are the cases where an incident is hardest to justify. ### Law firms and tax offices - **Typical setup:** Dragon Legal or Professional on Windows desktops or a Terminal Server, dictating into Word, RA-MICRO or DATEV. - **Priority:** Replace v14 and v15 seats and check BRAO and § 203 StGB obligations for any new cloud processor. See [Sprecho for law firms](https://sprecho.ai/solutions/legal). ### Medical practices and clinics - **Typical setup:** Older Dragon desktop seats next to Dragon Medical One, with patient data on every screen. - **Priority:** Health data falls under Art. 9 GDPR, so unpatched dictation PCs belong at the top of the risk register. See [Sprecho for healthcare](https://sprecho.ai/solutions/healthcare). ### Public administration - **Typical setup:** Centrally managed fleets with long Windows 11 rollouts and strict procurement. - **Priority:** A tool that installs on Windows 10 and 11 alike, deploys by MSI and is billed by invoice avoids a second migration later. See [Sprecho for public administration](https://sprecho.ai/solutions/government). ## Conclusion Your Dragon is not going to stop working overnight, but it is no longer safe by default: Windows 10 has been out of support since 14 October 2025, Dragon v14 since 1 September 2023, and Nuance ties every Windows 10 problem to a paid move to Windows 11 and Dragon v16, a product without a new major version since February 2023. If your users need offline recognition or Dragon's command grammar, upgrade to Dragon 16 on x64 Windows 11 hardware and budget for it. For everyone else, a continuously updated, EU-hosted tool that runs on Windows 10, Windows 11, Mac, Linux and mobile is the more future-proof choice. **Start your 14-day Sprecho trial without a credit card and pilot it on your Windows 10 and 11 machines this month**, or read the full [Sprecho vs Dragon comparison](https://sprecho.ai/compare/sprecho-vs-dragon) first. ## FAQ ### Does Dragon 15 work on Windows 11? Often it does, but Windows 11 is not Nuance's stated path for v15: Nuance positions Dragon v16 as the version optimized for Windows 11. Users report that Dragon 15.6 stopped dictating into anything except DragonPad after the Windows 11 24H2 update. If you depend on v15, test every Windows feature update in a pilot group first. ### Is Dragon 16 compatible with Windows 11 24H2? Dragon Professional 16 lists Windows 11 (64-bit) as a supported operating system and Nuance describes it as optimized for Windows 11. We found no Nuance statement specific to 24H2, and the breakage reports we found concern Dragon 15.6, not v16. Run a short pilot on a 24H2 machine before a fleet-wide update anyway. ### Is Dragon 14 still supported? No. Nuance ended all support for Dragon v14, in every edition and language, on 1 September 2023, and no further security patches or fixes are provided. Nuance itself warns that continued use may increase your exposure to security risks or affect your ability to meet compliance obligations. ### Can I keep Dragon on Windows 10? Technically yes: Nuance says Dragon v15 and v16 continue to be supported in Windows 10 environments. But if Nuance determines an issue is caused by Windows 10, you must upgrade to Windows 11 and Dragon v16. Windows 10 itself only receives security fixes through Microsoft's paid Extended Security Updates, at $61 per device in year one for organisations. ### Is unsupported dictation software a GDPR problem? It can be. Art. 32 GDPR requires security measures that reflect the state of the art, and dictation often contains health data under Art. 9 GDPR or client secrets protected by § 203 StGB. Running an unpatched operating system or unsupported software on that data is difficult to justify in your TOMs and in front of a data protection authority. ### Does Dragon run on ARM Windows or Copilot+ PCs? No. Dragon Professional does not support ARM-based Windows 11, and the Dragon 16.1 system requirements list only Intel or AMD processors. For the same reason Dragon cannot run in Parallels on Apple Silicon Macs, so check the CPU before you buy replacement laptops for dictation users. ### Does Sprecho support Windows 10? Yes. Sprecho requires Windows 10 or later, 64-bit, and the same installer covers Windows 10 and Windows 11, so you can switch dictation before your Windows 11 migration is complete. Recognition runs on Sprecho's servers in Germany, so an internet connection is required. Keep in mind that Windows 10 itself still needs ESU for security patches. ### Can IT deploy Sprecho via MDM? Yes, on the Enterprise plan. Admins download a signed MSI for Intune or Group Policy, a DMG for Jamf and a .deb for Linux from the Enterprise console, and devices enrol in the team when the user signs in, ideally via SSO. The device policy lets you turn off automatic updates or pin a version. --- ## https://sprecho.ai/blog/dragon-for-mac-discontinued-best-dictation-apple-silicon-2026 — Dragon for Mac Was Discontinued in 2018: The Best Dictation Software for Apple Silicon in 2026 > Dragon for Mac was discontinued in 2018 and won't run on Apple Silicon. Compare the 5 best Mac dictation apps for German professionals in 2026 (GDPR). *Published: 2026-10-05T12:00:30.873000Z* *Tags: dragon, mac, alternatives, gdpr, voice-dictation, comparison, germany* If you are looking for a **Dragon for Mac alternative** in 2026, the short answer is that there is no Dragon to go back to: Nuance discontinued Dragon Professional Individual for Mac on 22 October 2018 and stopped selling and updating it on the same day. Every Mac with an Apple Silicon chip was released after that date, so no version of Dragon has ever run natively on an M-series Mac. This guide explains what happened to Dragon Dictate and Dragon for Mac, why running the Windows version in Parallels does not solve the problem, what changed when Dragon Anywhere stopped selling in July 2026, and which **dictation software for Apple Silicon** fits German lawyers, doctors, tax advisors and consultants. We compare **5 alternatives** honestly, including where Dragon and the local-only Mac tools are still stronger. *Last verified: October 2026.* ## What happened to Dragon for Mac? **Dragon for Mac was discontinued on 22 October 2018.** Nuance's own advisory says Dragon Professional Individual for Mac was discontinued effective 10/22/2018, would "no longer be available for purchase", and that Nuance "will no longer provide updates for the Software" after that date ([Nuance advisory 27843](https://nuance.custhelp.com/app/answers/detail/a_id/27843/)). - **Dragon Dictate (up to version 4):** the original name of the Mac product, and still what many people type into Google today. - **Dragon for Mac 5 (2015):** the release that dropped the "Dictate" brand. - **Dragon Professional Individual for Mac 6 (2016):** the last Mac release, sold until October 2018. - **22 October 2018:** sales and updates end. Licences are perpetual, so an old installation may keep running, but nothing gets fixed for new macOS versions. - **Late 2020:** Apple begins the move to its own Apple Silicon chips. Dragon never received a native build for them. - **4 March 2022:** Microsoft completes the Nuance acquisition (about $19.7 bn) and invests in clinical AI (Dragon Copilot), not in a Mac comeback. - **1 July 2026:** Dragon Anywhere, the last Dragon dictation app you could buy for an iPhone or iPad, can no longer be bought or renewed. The Mac story ended eight years ago, and in 2026 the mobile story ended too. "Dragon" today means **Dragon Professional 16**, a perpetual, on-device desktop licence for Windows released on 28 February 2023, plus cloud subscriptions such as **Dragon Medical One** (healthcare, Microsoft Azure) and **Dragon Legal Anywhere** (law firms, Deutsche Telekom cloud), which also use Windows clients. ## Why do Mac users need a Dragon alternative in 2026? **Because no supported Dragon product runs on macOS, on Apple Silicon or on iOS any more.** If your firm works on MacBooks, Dragon is not something you can buy, install or keep up to date. - **No Mac app since 2018:** an eight-year-old Dragon 6 installation receives no fixes for new macOS releases, new microphones or new security requirements. - **No Apple Silicon build:** Dragon was never compiled for M-series chips. - **No mobile product:** Dragon Anywhere stopped selling and renewing on 1 July 2026, so iPhone dictation with Dragon is gone as well. - **Windows-only successors:** Dragon Professional 16 ($699.99 in the US, roughly 446 to 854 € netto at German resellers), Dragon Professional Anywhere and Dragon Medical One all need Windows. - **No major version since February 2023:** there is no Dragon Professional 17, so even the Windows line has gone more than three and a half years without a major release. - **Mixed fleets suffer most:** a partner on a MacBook and an assistant on a Windows PC cannot share one Dragon vocabulary, because only one of them can run Dragon at all. To be fair, Dragon still wins on several points, and you should know them before you switch: - **Richest command grammar:** "Select", "Correct That", "Insert Before", MouseGrid and custom commands have no exact equivalent in modern AI dictation tools. - **Fully offline on Windows:** Dragon Professional 16 recognises speech 100 % on-device, which is a real privacy strength. - **Three decades of Fachsprache:** Dragon's German legal and medical vocabularies are mature, and Dragon Medical One integrates deeply with PVS and KIS systems. ## Can you run Dragon on an Apple Silicon Mac with Parallels? **No, not on an M-series Mac.** Apple Silicon Macs can only virtualise the Arm version of Windows 11, and Microsoft's support page names Parallels Desktop as an authorised way to run the Arm versions of Windows 11 Pro and Enterprise on these Macs ([Microsoft support](https://support.microsoft.com/windows/options-for-using-windows-11-with-mac-computers-with-apple-m1-and-m2-chips-cd15fd62-9b34-4b78-b0bc-121baa3c568c)). Dragon Professional does not support Arm-based Windows 11, so the virtual machine has no supported Dragon to install. - **Apple Silicon (M1 to M5):** only Windows 11 on Arm runs in the VM, and Dragon Professional is not supported there. - **Intel Macs with Parallels:** an x86 Windows VM can run Dragon, but the virtualised microphone adds latency, which hurts recognition and makes long dictations tiring. - **Windows apps only:** Dragon inside the VM can only dictate into Windows applications in that VM, never into native macOS apps such as Mail, Pages, Safari or Word for Mac. - **Two systems to maintain:** you license, patch and secure a complete Windows installation just to keep one dictation tool alive. - **Ageing hardware:** the Intel-plus-Parallels route depends on Macs that are already several years old and will not be replaced by Intel models. The practical result: Parallels can keep an old Dragon workflow alive on an old Intel Mac, but it cannot give you Dragon in macOS, and it cannot give you Dragon on any Mac bought in the last few years. ## Dragon Anywhere is gone too: what changed on 1 July 2026? **Dragon Anywhere, Nuance's iOS and Android dictation app, can no longer be bought or renewed since 1 July 2026.** The App Store listing states: "As of July 1, 2026, Dragon Anywhere Mobile is no longer available for sale. Purchasing new subscriptions and renewing existing subscriptions is no longer possible." ([App Store listing](https://apps.apple.com/us/app/dragon-anywhere/id1024652126)) - **What it was:** a cloud dictation app for iPhone, iPad and Android at $14.99 per month or $149.99 per year. - **Last version:** 1.96.1, released in July 2026. - **What it means:** existing subscriptions run until their end date and cannot be extended. - **For Mac users specifically:** Dragon Anywhere was the last Nuance dictation product for individual professionals that ran on Apple hardware at all. With it gone, there is no Dragon left in your Apple setup. ## What does a Mac-based professional in Germany actually need? **A native Apple Silicon app that dictates into every macOS application, understands German Fachsprache and is run by an EU company you can sign a data processing agreement (AVV) with.** Everything else is a nice-to-have. - **Native Apple Silicon app:** no virtual machine and no workaround, but a build made for M-series Macs. - **Dictation into any app:** Mail, Outlook for Mac, Pages, Word for Mac, Safari and Chrome, including browser-based tools such as DATEV web applications and online forms. - **German Fachsprache:** legal, medical and tax vocabulary out of the box, plus a dictionary for your own names and terms. - **Windows and mobile for the rest of the team:** assistants on Windows PCs and partners on iPhones should use the same tool. - **One shared dictionary:** client names, court names and drug names entered once and available on every Mac and PC in the firm. - **EU company and EU hosting:** for § 203 StGB and GDPR Art. 9 work, the provider's jurisdiction matters as much as the server location. - **Predictable updates:** Apple ships a major macOS release every year, and your dictation tool has to keep up. ## Server location ≠ GDPR compliance: the CLOUD Act problem **A German data centre does not make a US provider safe for confidential work.** The US CLOUD Act (2018) lets US authorities compel US companies to hand over data regardless of where the server stands. | Setup | US jurisdiction? | GDPR-safe for confidential work? | |---|---|---| | US company + US servers | Yes | ❌ | | US company + EU servers | Yes (CLOUD Act) | ❌ | | EU company + EU servers | No | ✅ | - **Lawyers, doctors, tax advisors:** under § 203 StGB, disclosing client or patient secrets to a processor that a foreign authority can compel is a criminal-law question, not only a data protection one. - **Health data:** diagnoses dictated into an Arztbrief are special-category data under GDPR Art. 9, and voice recordings can be biometric data. - **Local tools sidestep the question:** Superwhisper in local mode, MacWhisper, and Apple Dictation on Apple Silicon for supported languages transcribe on your Mac. That is a genuine privacy advantage; the trade-off is features, platform coverage and paperwork (no AVV). - **Dragon's own cloud successors:** Dragon Medical One and Dragon Professional Anywhere run on Microsoft Azure. Microsoft is a US company, so the CLOUD Act applies even to the German-hosted variants. ## Top 5 Dragon for Mac alternatives compared **Sprecho is our pick for professionals and firms, Superwhisper and MacWhisper are strong for individual Mac users, Apple Dictation is the free fallback, and Wispr Flow suits English-first teams.** We weighted native Apple Silicon support, live dictation into any app, German vocabulary, team features and jurisdiction. - **Ranking logic:** built for German professionals on Macs, often in firms where some colleagues still use Windows. - **Disclosure:** Sprecho is our product. Every rival claim below matches what we publish on our comparison pages, and we say where rivals are better. ### 1. Sprecho — Best for professionals and mixed Mac/Windows teams **Sprecho** is a voice dictation app by Melo Designer GmbH (Lastrup, Germany) that types formatted text into any application while you hold a push-to-talk key. - **Native on every Mac:** separate builds for Apple Silicon and Intel, macOS 10.15 Catalina or later. Hold the **fn** key, speak, release, and the text lands in Mail, Pages, Word, Safari or any other app. - **Every other platform too:** native desktop apps for Windows and Linux plus apps for iOS and Android, with 3 devices per licence. - **German Fachsprache and your own words:** a personal dictionary with misspelling rules and auto-learn suggestions, and a **team-shared dictionary** on Pro and Enterprise that works the same on Mac and Windows. - **AI formatting:** punctuation, filler-word removal and a writing style per app (formal in Mail, short in Slack), plus snippets with variables such as {date} and {clipboard}. - **EU processor:** STRATO servers in Germany, GPU inference at Media Trooper (Germany/Netherlands), no US sub-processor in the dictation path, AVV available for Enterprise. - **Honest limits:** Sprecho needs an internet connection because recognition runs on its servers in Germany, not on your Mac. It also has no Dragon-style "Select" / "Correct That" / MouseGrid grammar and relies on AI formatting plus simple commands such as "new paragraph" and "scratch that". Details: [Sprecho vs Dragon compared](https://sprecho.ai/compare/sprecho-vs-dragon) and the guide to [installing Sprecho on a Mac](https://sprecho.ai/help/getting-started/install-sprecho-on-mac). ### 2. Superwhisper — Best for solo Mac users who want local processing **Superwhisper** is a Whisper-based dictation app from SuperUltra, Inc. (Canada) that runs speech recognition on your Mac by default. - **Strength:** fully on-device transcription on macOS by default, the most private setup for an individual; also available for Windows and iOS. - **Pricing:** $8.49 per month, $84.99 per year or $249.99 lifetime, billed in USD. - **Limits for firms:** no Linux or Android app, no shared team dictionary, no public AVV, and cloud post-processing uses your own API key with providers such as OpenAI or Anthropic. - **German:** Whisper handles German well, but there is no documented German Fachsprache tuning. Full breakdown: [Sprecho vs Superwhisper](https://sprecho.ai/compare/sprecho-vs-superwhisper). ### 3. MacWhisper — Best for transcribing recordings on a Mac **MacWhisper** is a Mac app by Good Snooze (Netherlands) that transcribes audio files locally with Whisper. - **Strength:** excellent local file transcription, rated "ausgezeichnet" by Macwelt, for €59 one-time. - **Not a live dictation tool:** system-wide dictation is a secondary, basic feature in the Gumroad version and absent from the App Store version. - **Limits:** Mac-only, single-user, no public AVV and an English-only interface. - **Best use:** interviews, recorded meetings and voice memos you transcribe afterwards. More: [Sprecho vs MacWhisper](https://sprecho.ai/compare/sprecho-vs-macwhisper) and our explainer on [transcription vs dictation with MacWhisper](https://sprecho.ai/blog/macwhisper-for-dictation-transcription-vs-dictation-explained-2026). ### 4. Apple Dictation — Best free fallback **Apple Dictation** is the speech-to-text feature built into macOS and iOS. - **Strength:** free, preinstalled, works in every Apple text field, and on Apple Silicon it runs on-device for supported languages. - **Privacy caveat:** where on-device processing is not available, audio goes to Apple's servers. Apple is a US company and offers no AVV for Dictation. - **Limits:** no custom vocabulary in standard Dictation, weaker accuracy on technical German, no team features, and no Windows, Linux or Android. - **Best use:** short messages and notes that contain no client or patient data. Comparison: [Sprecho vs Apple Dictation](https://sprecho.ai/compare/sprecho-vs-apple-dictation). ### 5. Wispr Flow — Best for English-first teams **Wispr Flow** is an AI dictation app from Wispr AI, Inc. (Delaware, USA) for Mac, Windows, iOS and Android. - **Strength:** fast, polished cloud dictation with mobile apps, snippets and team dictionaries. - **Pricing:** €15 per month, or €12 per month billed yearly (€144 per year). - **Limits for Germany:** primary processing on AWS in Virginia (US) with full CLOUD Act exposure, Flow Styles only in English, and no Linux build. - **Best use:** English-language teams without § 203 obligations. Comparison: [Sprecho vs Wispr Flow](https://sprecho.ai/compare/sprecho-vs-wispr-flow). ![Laptop and smartphone on a clean desk: one dictation setup across Mac, Windows and mobile](https://api.sprecho.ai/media/2026/10/1791196935-84c3c7a2.jpg) ## Feature comparison: Dragon for Mac alternatives side by side **Only Sprecho covers native Apple Silicon, Windows, Linux and both mobile platforms with an EU company behind it; the local Mac tools win on offline privacy.** | Feature | Sprecho | Dragon on a Mac | Superwhisper | MacWhisper | Apple Dictation | Wispr Flow | |---|---|---|---|---|---|---| | Native macOS app for Apple Silicon | ✅ | ❌ (discontinued 2018) | ✅ | ✅ | ✅ | ✅ | | Windows version for the rest of the firm | ✅ | ✅ (Professional 16, Windows only) | ✅ | ❌ | ❌ | ✅ | | Linux | ✅ | ❌ | ❌ | ❌ | ❌ | ❌ | | iOS / Android | ✅ / ✅ | ❌ (Dragon Anywhere not sold since 1 July 2026) | ✅ / ❌ | ❌ (no live dictation) | ✅ / ❌ | ✅ / ✅ | | Live dictation into any macOS app | ✅ | ❌ | ✅ | Basic (Gumroad version) | ✅ | ✅ | | German legal / medical vocabulary | ✅ Fachsprache + dictionary | ❌ on Mac (✅ on Windows) | ❌ | ❌ | ❌ | ❌ | | Shared team dictionary | ✅ (Pro, Enterprise) | ❌ | ❌ | ❌ | ❌ | ✅ (team plans) | | EU company + EU hosting | ✅ | ❌ (Microsoft, US) | Canadian company, local by default | Local on device | US company, on-device where supported | ❌ (US, AWS Virginia) | | Works offline | ❌ | ✅ (Windows desktop) | ✅ | ✅ | Partly | ❌ | | Price | €10.92/mo net | $699.99 one-time (Windows) | $8.49/mo | €59 one-time | Free | €15/mo | - **How to read "Dragon on a Mac":** the column shows what you can buy for a Mac in 2026, which is nothing native. Dragon Professional 16 runs only on Windows. - **Offline row:** if dictation must work with no network at all, a local tool is the honest answer; Sprecho always needs a connection. - **Team row:** only Sprecho and Wispr Flow share vocabulary across a team, and only Sprecho does it with an EU company and an AVV. ## Who benefits most from switching? **Firms where Macs and confidential data meet: law firms, medical practices, tax advisors and consultancies.** These are the readers who lost Dragon on the Mac first and who cannot simply fall back to a US cloud tool. ### Law firms (Kanzleien) ![Lawyer dictating at a desk in a German law office with legal books in the background](https://api.sprecho.ai/media/2026/10/1791196935-8eb003eb.jpg) - **Partners on MacBooks, staff on Windows:** one tool and one shared dictionary for court names, client names and standard clauses on both platforms. - **Schriftsatz boilerplate as snippets:** a spoken trigger inserts recurring paragraphs, signatures or addresses. - **§ 203 StGB and BRAO:** an EU processor with an AVV instead of a US provider. More on [Sprecho for law firms](https://sprecho.ai/solutions/legal). ### Doctors and practices - **Arztbriefe on the Mac:** dictate into the browser-based or Mac software you already use, with medical terms in your dictionary. - **GDPR Art. 9:** health data stays with a German GmbH on servers in Germany. See [Sprecho for healthcare](https://sprecho.ai/solutions/healthcare). - **Honest note:** Dragon Medical One's deep PVS/KIS integrations on Windows are more mature; Sprecho dictates into any text field instead of integrating with specific systems. ### Tax advisors and consultants - **DATEV and client mail:** dictate into DATEV web applications in the browser, into Outlook or Mail, with a formal style for email. - **Invoice in EUR with VAT:** a German invoice from a German GmbH. See [Sprecho for tax consultants](https://sprecho.ai/solutions/tax-consultant). ### Mixed Mac and Windows teams - **Same tool on every desk:** macOS, Windows and Linux desktops plus iOS and Android phones. - **Shared dictionary and snippets:** owners and admins manage shared entries, members keep personal ones; see the [shared team dictionary guide](https://sprecho.ai/help/teams-and-enterprise/shared-team-dictionary-and-snippets). - **Enterprise controls:** SSO/SAML, MDM deployment, audit-log export and payment by invoice. ### Pricing: Sprecho vs Dragon and the Mac alternatives | Product | Price | Model | Platforms | |---|---|---|---| | Sprecho Free | €0 (2,000 words/week) | Free forever, no card | Mac, Windows, Linux, iOS, Android | | Sprecho Pro | €10.92/mo net, or €9.24/mo net billed yearly (€110.82/yr) | Subscription | Mac, Windows, Linux, iOS, Android | | Sprecho Enterprise | €21.84/mo net, or €18.47/mo net billed yearly (€221.64/yr) | Subscription + SSO/SAML, MDM, audit-log export, invoice billing | Mac, Windows, Linux, iOS, Android | | Sprecho On-premise | On request | Customer-hosted (~50+ seats) | Mac, Windows, Linux | | Dragon Professional 16 | $699.99 one-time (US); ~446 to 854 € netto at German resellers | Perpetual licence | Windows only | | Superwhisper Pro | $8.49/mo or $84.99/yr; $249.99 lifetime | Subscription or lifetime | Mac, Windows, iOS | | MacWhisper | €59 one-time | Lifetime licence | Mac only | | Apple Dictation | Free | Bundled with macOS/iOS | Mac, iOS | | Wispr Flow Pro | €15/mo, or €12/mo billed yearly (€144/yr) | Subscription | Mac, Windows, iOS, Android | - **VAT:** Sprecho's net prices plus 19 % VAT are €12.99 per month, or €10.99 per month billed yearly (€131.88 per year), invoiced by a German GmbH. - **Three years, one seat:** Sprecho Pro yearly costs 3 × €110.82 = €332.46 net. For a Mac user the Dragon comparison is moot, because no Dragon price buys a Mac app. - **Try before you decide:** a 14-day Pro trial without a card, and a 14-day Enterprise test for teams. All plans on the [pricing page](https://sprecho.ai/pricing). ## How do you switch from Dragon on a Mac to Sprecho? **Carry over your vocabulary, install the Apple Silicon build, grant two macOS permissions and rebuild your auto-texts as snippets.** Most people are dictating within minutes. 1. **Save your word list:** if you still have a Windows Dragon profile, open Tools > Vocabulary Center > "Export custom word and phrase list" (TXT or XML). From an old Mac installation, copy your custom words into a text file. 2. **Bring it into Sprecho:** there is no self-service bulk import yet, so send us your exported list and we convert it, or add the entries yourself in the [dictionary](https://sprecho.ai/help/formatting-and-ai/teach-sprecho-your-words-with-the-dictionary). 3. **Install the right build:** the download page picks Apple Silicon or Intel automatically. 4. **Grant Microphone and Accessibility:** the second permission lets Sprecho type into other apps; restart the app after granting it. 5. **Choose your key:** hold fn (the default) for short dictations, or switch to lock mode for long letters. 6. **Rebuild auto-texts as snippets:** signatures, addresses and standard paragraphs become [voice snippets](https://sprecho.ai/help/formatting-and-ai/snippets-voice-shortcuts-for-text-you-repeat). 7. **Test in your real apps:** Mail, Word, Pages and your browser-based tools during the 14-day trial. How your Dragon habits translate: - **"Scratch That":** say "scratch that" to remove the phrase you just spoke. - **"New Paragraph" / "New Line":** work the same way in Sprecho. - **Punctuation by voice:** comma, period, question mark, quotes, brackets and the section sign (§), each after a short pause of about half a second; see [voice commands](https://sprecho.ai/help/formatting-and-ai/dictating-voice-commands-reliably). - **"Select" / "Correct That" / MouseGrid:** no equivalent. You correct with the keyboard, and AI formatting removes most of the need for explicit commands. ## How to choose the right Dragon for Mac replacement **Decide by who dictates, on which devices and with which data.** - **A firm with confidential data and mixed devices:** Sprecho, because it covers Mac, Windows, Linux and mobile with one EU contract and a shared dictionary. - **A single person who must stay offline:** Superwhisper in local mode, accepting no team features and no AVV. - **Mostly recordings, not live typing:** MacWhisper. - **Short private notes only:** Apple Dictation is free and good enough. - **English-only team without professional secrecy:** Wispr Flow. - **Heavy Dragon command users on Windows:** stay on Dragon Professional 16 for now if you depend on Select/Correct grammar, and plan the Mac side separately. ## Conclusion Dragon for Mac has been dead since 22 October 2018, Parallels cannot bring it back on Apple Silicon, and since 1 July 2026 Dragon Anywhere is gone from iPhones too. Local tools such as Superwhisper and MacWhisper are excellent for individuals who want offline processing, but a German law firm, practice or tax office needs native Mac and Windows apps, mobile, a shared dictionary and an EU processor with an AVV. That combination is what Sprecho is built for. **Start your 14-day free Pro trial, no card required**, on the [Sprecho pricing page](https://sprecho.ai/pricing), or read our full guide to the [best Dragon alternatives in 2026](https://sprecho.ai/blog/best-dragon-alternatives-2026). ## FAQ ### Is Dragon available for Mac in 2026? No. Nuance discontinued Dragon Professional Individual for Mac on 22 October 2018 and stopped both sales and updates on that date. The only desktop Dragon sold today is Dragon Professional 16 for Windows, and Dragon Anywhere for iPhone and iPad can no longer be bought or renewed since 1 July 2026. ### Can I run Dragon on an M-series Mac with Parallels? Not in a supported way. Apple Silicon Macs (M1 to M5) can only run the Arm version of Windows 11 in Parallels, and Dragon Professional does not support Arm-based Windows 11. On older Intel Macs Dragon can run in an x86 Windows VM, but with microphone latency and only into Windows apps inside the VM, never into Mail, Pages or Safari. ### What replaced Dragon Dictate? Dragon Dictate was renamed Dragon for Mac with version 5 in 2015 and became Dragon Professional Individual for Mac with version 6, which Nuance discontinued in October 2018 without a Mac successor. Today the closest replacements are native Apple Silicon dictation apps: Sprecho for system-wide professional dictation, or Superwhisper and MacWhisper for local use by individuals. ### What is the best Dragon alternative for Mac in Germany? For professionals and firms it is Sprecho: native builds for Apple Silicon and Intel Macs, plus Windows, Linux, iOS and Android, German Fachsprache, a shared team dictionary and an AVV from Melo Designer GmbH with servers in Germany. If you are an individual who must dictate fully offline, Superwhisper in local mode is the strongest option, but it has no AVV and no team dictionary. ### Is Apple Dictation GDPR-compliant for § 203 StGB work? Apple Dictation runs on-device on Apple Silicon for supported languages, which is good for privacy, but otherwise audio is sent to Apple's servers, and Apple is a US company subject to the CLOUD Act. Apple offers no AVV for Dictation, which lawyers, doctors and tax advisors bound by § 203 StGB, or anyone dictating health data under GDPR Art. 9, would normally need. Use it only for notes without client or patient data. ### Does Sprecho work in Word for Mac and Apple Mail? Yes. Sprecho types into whichever app has focus, including Word for Mac, Apple Mail, Outlook, Pages, Safari, Chrome and browser-based tools, once you grant macOS the Microphone and Accessibility permissions. It needs an internet connection because recognition and formatting run on Sprecho's servers in Germany, not on your Mac. ### Does Sprecho share a dictionary between Mac and Windows colleagues? Yes, on Pro and Enterprise. Entries in the team-shared dictionary and shared snippets apply to every team member on every platform, whether they dictate on a Mac, a Windows PC or Linux. Owners and admins manage shared entries, while each person's personal entries stay private. ### How does Sprecho's price compare with Dragon? Dragon Professional 16 costs $699.99 one-time in the US and roughly 446 to 854 € netto at German resellers, but it runs only on Windows, so no Dragon price buys you a Mac app. Sprecho Pro costs €10.92 net per month, or €9.24 net per month billed yearly (€110.82 per year), which is €12.99 or €10.99 including 19 % VAT. There is also a free plan with 2,000 words per week and a 14-day Pro trial without a card. --- ## https://sprecho.ai/blog/dragon-voice-commands-in-sprecho-2026 — Dragon Voice Commands in Sprecho: What Works, What Doesn't, and What Replaces Them (2026) > Dragon voice commands list mapped to Sprecho: punctuation, Scratch That, Select, Correct, MouseGrid and custom commands in English and German, verified 2026. *Published: 2026-10-05T12:00:26.730000Z* *Tags: dragon, voice-dictation, migration, comparison, gdpr, germany, windows* **Dragon voice commands list** is one of the first things long-time Dragon users search for when they try a modern dictation app, because twenty years of muscle memory ("comma", "new paragraph", "scratch that") do not disappear overnight. This cheat sheet maps every Dragon command you are likely to use onto Sprecho: which ones work with the same words, which ones do not exist, and what replaces them. *Last verified: October 2026.* - **Punctuation and layout:** almost all of Dragon's everyday punctuation words work in Sprecho, in English and German, with a few wording differences listed below. - **Correction:** "Scratch that" / "Streich das" works, and fillers and false starts disappear on their own. Dragon's Select, Correct, Cap and Insert grammar does not exist in Sprecho. - **App and mouse control:** "Open Word", "MouseGrid" and "Click