WhisperWeb
WhisperWeb is a free AI transcription tool that turns local audio and video files into editable text and subtitles. It runs the Whisper speech recognition model in your browser, making it useful for interview excerpts, lecture recordings, voice notes, and short podcast segments.
Recordings and transcripts are processed on your device, rather than uploaded to a cloud transcription service. Choose a file or record audio in the browser, select the spoken language or automatic detection, and review the result against the recording. You can correct words, adjust subtitle start and end times, and export TXT, SRT, or VTT files without creating an account.
WhisperWeb supports common formats including MP3, WAV, M4A, MP4, WebM, and MOV. It also converts existing SRT and VTT subtitle files between formats or extracts their text. For a YouTube video you manage, you can generate a subtitle file locally and upload it through YouTube Studio. WhisperWeb itself needs a saved video file; it does not transcribe directly from a YouTube link.
Transcription preserves the spoken language. You can also translate speech into English, but other translation targets and dubbing are not supported.
Each file is limited to 200 MB and 20 minutes. Longer recordings need to be split before processing. The first session downloads a speech model, and later sessions can reuse it while the browser cache remains available. Tiny, base, and small models give you different download and memory requirements; processing speed depends on your hardware.
Desktop Chrome on macOS is the documented tested setup. Mobile transcription has not been verified by the developer, so a phone-friendly page should not be taken as a guarantee that your phone can handle recognition. Check names, figures, and technical vocabulary before using a transcript.
Export your results before closing the task because there is no cloud recovery. Local processing does not mean the website makes no network requests: loading the site and model downloads requires a connection, and anonymous usage counts are optional. Transcription, transcript review, and subtitle exports are free.
Alternatives to WhisperWeb
- SpotScribe: Transcribes podcasts and uploaded recordings, with AI summaries and questions about the content.
- Speak AI: Adds searchable recording libraries, meeting capture, summaries, and analysis across multiple conversations.
- Cleanvoice: Combines transcription with podcast audio cleanup, filler removal, and show notes.

TRY IT