What Happened to Windows Speech Recognition? Where It Went and What Replaces It

If you use Windows Speech Recognition and recently found it missing from your system, you have not done anything wrong. Microsoft confirmed that Windows Speech Recognition removed status applied to Windows 11 from September 2024 onwards — and for many users the change arrived with no warning. WSR had been part of Windows since 2006, and some people had spent years building up trained voice profiles and custom vocabulary lists they relied on every single day. Finding it gone is genuinely disorienting.

This article explains exactly what happened, why Microsoft made the change, what Windows Voice Access does and does not offer as the WSR replacement, and what options exist for the many users who need more than either tool ever provided.

For users who specifically need voice typing inside Microsoft Word, our guide on how to dictate in Word covers the current built-in option and its Microsoft 365 requirements.

For a practical look at what Windows 11 offers today, our guide to free speech to text on Windows 11 tests every current built-in option side by side.

Windows Speech Recognition Removed: The Official Timeline

Windows Speech Recognition arrived with Windows Vista in 2006. For nearly two decades it gave Windows users a locally processed way to dictate documents, navigate the desktop, and issue voice commands entirely without an internet connection. It was also a meaningful accessibility tool for people with repetitive strain injuries and conditions that make keyboard use difficult or painful.

In December 2023, Microsoft officially added WSR to its list of deprecated Windows features on Microsoft Learn, signalling that active development had ended. The listing stated plainly that WSR was being replaced by Voice Access. At that point the feature was deprecated but not yet removed — it still ran on affected machines.

From September 2024, the change began to be enforced. Windows 11 versions 22H2 and 23H2 had Windows Speech Recognition deprecated and removed as part of a system update, replaced automatically by Voice Access. Windows 11 version 21H2 — the original release — was not immediately affected, but as machines update to newer builds, WSR disappears. Windows 10 still includes WSR, but Windows 10 itself reached end of support in October 2025 and no longer receives security updates. Relying on it to keep WSR running is a diminishing option.

In short: if you are on a current Windows 11 machine and WSR is gone, that is deliberate and permanent.

What Microsoft Gave You Instead: Windows Voice Access

Voice Access is Microsoft’s modern answer to the question of how you control a Windows PC by voice. It first appeared in Windows 11 in 2022 and became the default built-in voice control system from September 2024 onwards. You can enable it via Settings, then Accessibility, then Speech — a small toolbar appears at the top of the screen when it is active.

The core capability is familiar: dictate text, click buttons, open applications, scroll pages, and navigate menus using spoken commands. Microsoft built Voice Access on more recent neural speech models than the acoustic approach WSR used, which generally produces better out-of-the-box accuracy on modern hardware, particularly for navigation commands.

Where Voice Access differs is in scope and flexibility. At launch it supported only six English locales: US, UK, India, New Zealand, Canada, and Australia. Subsequent updates have expanded this, but coverage remains narrower than WSR, which could work with any Windows language pack you had installed. More critically for long-time users, Voice Access does not import existing voice training data, custom vocabulary lists, or the macro-style command shortcuts that power users had assembled over years. You are starting fresh, learning a new command vocabulary, with no path to carry your history forward.

Many users who made the switch have also reported that Voice Access handles text entry differently from WSR, and the transition requires relearning habits that had become muscle memory. Forum threads on Microsoft Q&A are filled with users asking whether there is a way to recover their WSR profile. There is not.

What You Actually Lost — and Why the Frustration Is Justified

The friction showing up in support communities is worth taking seriously, because what WSR users lost is not just a feature — it is years of accumulated work. WSR improved over time by learning from you. The more you used it, corrected its mistakes, and added custom vocabulary, the better it adapted to your voice and speaking patterns. Many users had profiles running continuously since Windows 7 or Windows 10, representing hundreds of hours of implicit training.

WSR also supported custom commands built around macros. You could speak a short phrase and have it trigger a full text block, a keyboard shortcut sequence, or an application action. For legal professionals, medical secretaries, journalists, and writers managing large dictation workloads, those macros were not a nice-to-have — they were the core of the workflow.

Voice Access does not replicate this. The command model is different, the dictation behaviour is differat, and the macro capability WSR power users depended on is simply absent. Windows dictation changed in a way that Microsoft considers an improvement overall — and for users who primarily need to control their PC without a keyboard, Voice Access probably is better. For pure dictation productivity, the gap is real and the workarounds are clumsy.

If you are currently hitting errors with voice input in Windows 11, the guide to Windows 11 voice typing not working covers the most common faults and how to fix them. For a broader look at what tools are worth using today, the best dictation software for Windows in 2026 gives an honest, up-to-date comparison of your options.

The Ceiling of Transcription — and the Upgrade That Goes Further

WSR’s removal surfaces something that was always true of transcription tools: their job ends the moment they have captured your words. Whether it is WSR or Voice Access, what happens next — editing for tone, restructuring sentences, correcting punctuation, adapting the register for an email versus a formal report — is still entirely yours to manage. The transcript is the beginning of the work, not the end of it.

That gap is most visible when you are producing the kind of content that fills most working days: replies to clients, internal updates, LinkedIn posts, messages in Teams or Slack, meeting follow-ups. You need output that is ready to send, not a rough transcript that still needs fifteen minutes of editing before it is usable.

This is exactly the problem that think-to-text addresses. Genie 007’s approach is distinct from both WSR and Voice Access: instead of capturing what you say literally, it interprets what you mean and produces what you need — formatted, appropriately toned, and placed directly in whatever app or text field you have open on Windows.

A practical example makes this concrete. You are in Outlook and need to tell a client their contract amendment is being finalised and will arrive by the end of the week. With a transcription tool, you dictate the reply sentence by sentence, clean in punctuation and phrasing, adjust the tone, and then send. With Genie 007 on Windows, you press a keyboard shortcut inside Outlook, say something like “tell them the amendment is ready and they will receive it by Friday”, and a complete, professional reply appears in the field. No separate panel to manage, no tab switching, no editing pass.

That is Genie Mode — and it works inside Gmail, Outlook, Teams, LinkedIn, Slack, Notion, Google Docs, and essentially every other application and website you use on Windows. For users who want straightforward fast dictation without AI rwriting, Genie 007 also offers Voice Typing mode: automatic punctuation, no dictation commands to memorise, 140 or more languages supported, and live translation if you need to write in a different language from the one you are speaking in.

There is no training period and no custom vocabulary to rebuild. The AI works from context and intent from day one. If you want to see both modes in action, the Genie 007 for Windows page covers setup and compatibility with the apps you already use. Pricing starts at £5 per month and there is a free trial with no credit card required.

Frequently Asked Questions

Is Windows Speech Recognition still available on Windows 11?

No. Microsoft deprecated Windows Speech Recognition in December 2023 and removed it from Windows 11 versions 22H2 and later from September 2024. If you are running a current Windows 11 machine, WSR is no longer present. Windows Voice Access is the built-in replacement and can be enabled via Settings, then Accessibility, then Speech.

What replaced Windows Speech Recognition in Windows 11?

Microsoft replaced Windows Speech Recognition with Windows Voice Access, which uses newer AI-based speech models for improved accuracy and hands-free PC control. Voice Access was first introduced in 2022 and became the standard for Windows 11 22H2 and later from September 2024. It supports dictation and full voice navigation but does not carry over WSR voice training profiles or custom commands from previous installations.

Can I still use Windows Speech Recognition on Windows 10?

Yes — WSR remains available on Windows 10. However, Windows 10 reached end of support in October 2025 and no longer receives security updates or patches. Remaining on Windows 10 solely to retain WSR is not a practical long-term strategy, and most users upgrading to Windows 11 will need to adapt to Voice Access or choose a dedicated voice tool as their WSR replacement.

Does Windows Voice Access work in all languages?

Voice Access launched with support for six English locales and has expanded since, but its language coverage remains more limited than WSR, which worked with any Windows language pack installed on the machine. If you regularly dictate in a language other than English, Voice Access may not meet your needs. AI-based voice tools with broader multilingual support — including live translation across 140 or more languages — are worth exploring as an alternative.

---

Try Genie 007 Free

WSR’s removal is a genuine opportunity to move to something better, not just an equivalent swap. If you want voice input that produces finished, ready-to-send output rather than raw transcription that still needs editing, Genie 007 is built for exactly that workflow.

Pownload Genie 007 free â² available for Windows, MacK, mobile and as a browser extension. No credit card required.

Written by Bill Kiani, founder of Genie 007.