TXT24 — Video to Text | PDF to Text
EVERY WORD, IN ONE PLACE

Turn recorded speech into a transcript you can use.

Give meetings, interviews, lessons and voice notes a readable home. Keep the recording nearby while you review the words.

Start with a recording

Upload MP3, WAV, M4A, MP4 or WebM, or record audio from your microphone after choosing Start recording.

Review the words

Edit transcript text and available segment times. Speaker labels can be renamed throughout the transcript.

Export for the next step

Download readable documents or SRT/VTT captions. Custom caption splitting uses matching word timestamps.

A CLEAR PATH

How it works

  1. Upload a recording or save a microphone recording to your library.
  2. Select its audio language, or leave automatic detection enabled.
  3. Choose speaker recognition when a compatible model is available, then submit transcription.
  4. Open the completed transcript, review it and save your edits before exporting.

Timing that follows your source

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.

Provider availability is visible

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.

A few useful answers

Does language selection translate the recording?

No. Audio language is a recognition hint. Translation is a separate action in AI Studio.

Can I write a transcript myself?

Yes. You can type or paste text, save it in the workspace and export a readable transcript without automatic speech recognition.