Skip to content
FileTinker

Video to Text

Extract the speech from a video as editable text. Drop in an MP4, WebM or other video; the audio track is decoded in your browser and transcribed by a local AI model — nothing is uploaded, so even a long lecture recording or meeting export stays private and costs nothing. Download the result as plain text or as SRT/VTT subtitles.

First use downloads a 77 MB speech model, then your browser caches it.

How to convert video to text

  1. Drop in a video file — MP4, WebM and most desktop formats work.
  2. Choose the model size and spoken language, then click Transcribe.
  3. Copy the transcript or download it as TXT, SRT or VTT.

About this video-to-text converter

This converter runs OpenAI's open-source Whisper model in the browser via WebAssembly. The first run downloads your chosen model size (41–249 MB, then cached); after that, transcription is a purely local computation. It's the same technology transcription services charge monthly for — minus the upload.

Video transcription shines for meetings, lectures, interviews and screen recordings: search the text instead of scrubbing the timeline, quote people accurately, or turn a talk into an article. Use auto-detect for mixed-language footage, or set the language explicitly for the best accuracy on single-language recordings.

Frequently asked questions

Which video formats work?

Whatever your browser can play: MP4 works in every modern desktop browser; WebM works in Chrome, Edge and Firefox; MOV and others depend on the codecs your system ships. Only the audio track is read — the pictures are ignored entirely. If a video won't decode, extract its audio first (this site has a dedicated extract-audio tool) and transcribe the MP3.

Is the video uploaded to a server?

No. The file is opened locally: your browser decodes the soundtrack and the speech model runs on your own device. That matters for meetings and lectures — you can transcribe confidential footage without it touching anyone's cloud.

Can I get subtitles instead of plain text?

Yes — the same transcription exports as SRT or WebVTT with timestamps, ready to load into a player or editor, or to publish alongside the video. Use the format buttons above the result to switch.

How long can the video be?

Up to 300 MB of file — that comfortably covers an hour or two of typical MP3 audio, but compressed video reaches the cap much sooner. For a long video, extract the audio track first (see the extract-audio tool) and transcribe that instead. Processing happens on your machine, so long files take a while on slower devices — the progress bar shows how far along it is, minute by minute.