Interview workflow

Turn an interview into editable timed text

Create a local first-pass transcript for journalism, documentary editing, qualitative research or content production while keeping the source recording on the device.

Quick answer

Whisper Online downloads a whisper.cpp model, processes the selected media locally, lets you correct individual timed words, and exports the result without uploading the media file.

Create a transcriptMedia stays on your device

Step by step

The practical workflow

  1. 1

    Select the interview recording

    Choose the audio or video file directly from local storage.

  2. 2

    Generate a timed draft

    Run Whisper in the browser and wait for the token-level transcript.

  3. 3

    Check quotes carefully

    Correct words, names and time ranges against the original recording.

  4. 4

    Export the reviewed material

    Download TXT for reading, JSON for time-aware workflows or subtitles for an edited video.

A first pass for journalists and researchers

Automatic transcription can accelerate review, but it should not be treated as a verified quotation. Listen back to consequential statements and correct the transcript before publication or analysis.

Individual timing makes it easier to keep edited words aligned with the source and provides structured data for custom review tools.

Privacy and informed handling still matter

Local processing avoids transferring the interview to this site's server. That reduces one data-sharing step, but the user remains responsible for consent, storage security, access control and any legal or ethical obligations applying to the recording.

  • Obtain appropriate consent before recording or transcribing.
  • Verify quotations against the original audio.
  • Do not assume automatic text identifies speakers correctly.
  • Store downloaded transcripts according to the sensitivity of the material.

Documentary and video interview captions

For an edited interview, export SRT and import it into the video editor. Caption grouping can be adjusted for line length and duration without changing the corrected word-level transcript.

Why use local transcription?

No media upload or account
Editable word-level timing
SRT, VTT, TXT, and JSON export

FAQ

Common questions

Is an automatic transcript accurate enough to quote?

Treat it as a draft. Verify important quotations, names and numbers against the recording.

Does it identify the interviewer and interviewee?

No. Speaker diarization and speaker labels are not currently included.

Can I keep exact word timings?

Yes. JSON export includes every edited token's individual start and end time.

Is the recording stored by the site?

No. The recording is processed locally in the browser and is not uploaded to this application.

Related guides