Start with a recording
Upload MP3, WAV, M4A, MP4 or WebM, or record audio from your microphone after choosing Start recording.
Sign in Give meetings, interviews, lessons and voice notes a readable home. Keep the recording nearby while you review the words.
Upload MP3, WAV, M4A, MP4 or WebM, or record audio from your microphone after choosing Start recording.
Edit transcript text and available segment times. Speaker labels can be renamed throughout the transcript.
Download readable documents or SRT/VTT captions. Custom caption splitting uses matching word timestamps.
Timed transcripts link segments to the recording so you can check a line in context. When the provider returns word timestamps that still match the text, subtitle formatting can group words into readable cues.
Text or timing edits can make word-level alignment stale. In that case, exports preserve the saved segment boundaries rather than inventing new word times.
Speech recognition is an external processing step. The workspace explains when speech credentials or a compatible speaker-recognition model are missing. Recording quality, language and overlapping speech can affect the result, so review important wording before sharing.
No. Audio language is a recognition hint. Translation is a separate action in AI Studio.
Yes. You can type or paste text, save it in the workspace and export a readable transcript without automatic speech recognition.