Windows Speech to Text Registry Path in Regedit
This guide outlines the specific Windows Registry paths that control
speech-to-text, dictation, and speech recognition settings in Windows.
In modern Windows operating systems, speech settings are divided between
the legacy Speech API (SAPI) paths and the modern OneCore speech engine
paths, located within both the HKEY_CURRENT_USER and
HKEY_LOCAL_MACHINE hives.
Modern Speech-to-Text and Dictation Registry Paths
Windows 10 and Windows 11 use the OneCore engine for dictation, voice typing, and modern speech recognition.
User-Specific OneCore Speech Settings:
HKEY_CURRENT_USER\Software\Microsoft\Speech_OneCoreSystem-Wide OneCore Speech Engines and Voices:
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Speech_OneCoreOnline Speech Recognition (Cloud Dictation Toggle):
HKEY_CURRENT_USER\Software\Microsoft\Speech_OneCore\Settings\OnlineSpeechPrivacy- Within this key, the
HasAcceptedDWORD value controls whether cloud-based speech recognition and dictation are enabled (1) or disabled (0).
- Within this key, the
Legacy Speech Recognition (SAPI) Registry Paths
The classic Windows Speech Recognition tool (found in the legacy Control Panel) stores its profile data and configuration under the standard Speech keys:
- User Profiles and Configuration:
HKEY_CURRENT_USER\Software\Microsoft\Speech...\Speech\Recognizers: Contains preferences and available speech recognizers for the active user....\Speech\CurrentUserLexicon: Contains customized user vocabulary and phonetic settings.
- System-Wide Recognizer Components:
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Speech...\Speech\Recognizers\Tokens: Contains registry registrations for all installed speech recognition engines on the machine.
How to Access and Modify the Paths
- Press
Win + Rto open the Run dialog. - Type
regeditand pressEnterto open the Registry Editor. - Paste any of the target paths above directly into the address bar at
the top of the Registry Editor window and press
Enterto jump straight to the speech configuration keys.