ChatGPT Launches Live Dictation, Turning Speech Into Text as You Talk
Dictation in ChatGPT is getting faster and more reliable. With Live Dictation, your speech turns into text as you talk, so you spend less time waiting for your message to finish transcribing. Yap away!
Context
A faster, live transcription path inside ChatGPT is a small change in the interface, but it is one more step toward talking to the assistant as the default, not typing.
The ChatGPT account posted that dictation in ChatGPT is getting faster and more reliable, and that with Live Dictation, speech turns into text as you talk, so people spend less time waiting for a message to finish transcribing. The post is an official product note and does not say which plans or platforms get the feature first.
Deep dive
ChatGPT's official account said dictation in ChatGPT is getting faster and more reliable, and that Live Dictation turns speech into text as you talk, so users spend less time waiting for a message to finish transcribing. A short video is attached. Those are first hand product claims from OpenAI's consumer account.
The post does not name languages, devices, or plans, and it does not report an accuracy figure. 'Faster and more reliable' is the company's own comparison with the prior dictation path, not an independent test. The useful change, as described, is streaming text rather than waiting for a full utterance to be transcribed at the end.
This is an incremental interface update, not a new model. It matters for daily use of ChatGPT, and it sits beside other recent packaging of the assistant, such as the Sales experience already on the Nov 17 deck. It should not be read as a speech model launch.
Because the lead is official, the feature name and the live transcription behavior are confirmed as OpenAI's announcement. Availability details will have to come from the app or a later post.
What to expect next
The ChatGPT app and later help pages should show which accounts and devices get Live Dictation, and in which languages. Watch for quality complaints or a later accuracy note. If the feature is limited to certain plans, that belongs here. This story should stay a product update, not a model release, unless OpenAI later ties it to a new speech system by name.
