Subtitle Generator
Generate SRT/VTT subtitles from video or audio with Whisper, locally.
How it works
- 1Select a video or audio file — it is decoded inside your browser, never uploaded.
- 2The audio track is extracted and resampled to 16 kHz mono on your device.
- 3Whisper transcribes with timestamps (the same ~40 MB model as Speech to Text, cached once).
- 4Review the segments, then download SRT or VTT, or copy them to the clipboard.
FAQ
Are my videos uploaded to a server?
No. Audio decoding and transcription run entirely inside your browser. Only the audio track is processed — video frames are discarded.
Which subtitle formats are supported?
SRT and WebVTT (VTT), plus a plain-text transcript with timestamps. SRT works with most video players and editors.
Which languages are supported?
Whisper is multilingual and auto-detects the spoken language. English works best; with the tiny model, Chinese and other non-European languages come out rough — sometimes translated or paraphrased — because the model itself is small.
How long does it take?
It runs on WASM — expect roughly real-time length (a 5-minute video takes a few minutes depending on your device).
Can I fix subtitle timings?
The SRT/VTT output is plain text — edit it in any subtitle editor or text editor after download. Cue boundaries come from Whisper's timestamps.