Turn any video into timed subtitles.

Yes — completely free. No sign up, no watermark, and no usage limits. The transcription runs on your own device, so there are no server costs to pass on.

No. The speech recognition model is downloaded to your browser and runs there. Your video never leaves your device, and nothing is stored anywhere.

Because the transcription happens on your device, the speech recognition model has to be downloaded once — roughly 170 MB. It is then cached by your browser, so every later visit starts instantly with no download at all.

SRT and WebVTT, which cover essentially every video editor and player, plus JSON if you want the raw word-level data with timings and confidence scores.

The model is multilingual and handles most major languages, though it is strongest on English. Accuracy depends a lot on audio quality — clear speech with little background noise transcribes far better than a noisy room.

Accurate enough to caption with, and honest about the difference. Timings start as estimates derived from the speech recogniser, which works at roughly one-second granularity. Precise word-level timing is a separate, optional step you can opt into.