Convert video to text: drop an MP4, MOV, MKV or WebM and the speech in it comes back as a transcript, with Whisper running on your own machine. The audio track is pulled out and decoded locally, so a recorded meeting, a lecture or a client call never leaves your device. Read and correct the text as it appears, then download it as plain .txt; for captions, switch Output to Subtitles before you run it and save timestamped .srt or .vtt instead.
How it works
- Drop the video file. Its audio track is decoded on your device; nothing is sent anywhere.
- Pick the spoken language and a model: Fast (≈41 MB) or Accurate (≈77 MB), downloaded once and then cached.
- Press Transcribe and the text appears live. Download it as .txt, or set Output to Subtitles first for line-timed .srt / .vtt.
Questions
- How do I convert a video to text for free?
- Drop the video on this page, pick the language it is spoken in, and press Transcribe. There is no account and no daily minute allowance, because the transcription runs on your own computer instead of a server. The first run downloads the model once; after that it is cached.
- Can it transcribe a YouTube video?
- Only from a file on your device: a web page cannot fetch a video from YouTube for you. If it is your own video, download it from YouTube Studio first, then drop the file here.
- Does my file get uploaded?
- No. The models and all processing run inside your browser on your own device. You can verify this yourself: open your browser’s DevTools, watch the Network tab, and process a file: there are no uploads. After the first visit most tools also work fully offline.
- How accurate is it?
- The models are OpenAI’s Whisper (tiny and base tiers). On clear speech they are very usable; heavy accents, crosstalk and noisy rooms will show errors. The Accurate tier is noticeably better on difficult audio.
- Which languages work?
- Fifteen, picked in the Language setting: English, Spanish, French, German, Italian, Portuguese, Dutch, Russian, Polish, Turkish, Japanese, Korean, Chinese, Arabic and Hindi. The decoder is conditioned on the one you pick, so it transcribes that language rather than approximating it in English — which also means picking the wrong one returns confident nonsense. The weights understand 99 languages; these are the ones we measured as good enough to offer.
- How long does it take?
- Transcription runs on your processor, so speed depends on the machine: expect around a quarter of the recording’s length, and less on the Fast model than on Accurate.