What we know
OpenAI has launched a Live Dictation feature for ChatGPT that transcribes speech in real time as users speak, instead of waiting until they finish talking. This update is currently being rolled out to users, enabling faster and more seamless voice input. ChatGPT can now transcribe audio prompts live, capturing speech as it happens. The Live Dictation feature is in the process of rolling out to users.
Why it matters
Before this update, ChatGPT required users to complete their spoken input before starting transcription, which could slow down interactions. Live Dictation allows for immediate transcription, making voice conversations with ChatGPT more natural and efficient. This improvement supports hands-free use and enhances accessibility, aligning with OpenAI’s goal to make ChatGPT easier and faster to use.
What is still unknown
Details about the accuracy of the live transcription, which languages are supported, and when the feature will be widely available remain unclear. No independent tests of transcription quality or performance have been reported.

