How do I transcribe a recording with Speech to Text?

Speech to Text turns audio into written text. For translating and dubbing a recording, use Interpretation instead.

1. Open Speech to Text and choose the file-upload mode. Select the recording you want to transcribe.
2. Wait for the file to be read. If it contains multiple audio tracks, select the intended track. Choose an available language or automatic detection.
3. Check the displayed cost and start transcription. When complete, review and edit the transcript in that project.
4. Use the project’s download menu and choose an available export format. You can reopen a saved project through History.

For an upload, transcription or export problem, tell Support the step, time, file type and visible error. Do not send the recording or full transcript unless requested. Real-time transcription is a separate mode; a file-upload problem does not establish a problem with that mode.

You can upload a video with audio without extracting the sound first. If it has several audio tracks, usage is based on the selected track rather than the length of the video picture. Reopen a submitted file transcription from History. The website can export a completed transcript as TXT, PDF, DOCX, JSON or HTML. For time-coded SRT subtitles, see the public transcription API download guide.