OpenSubs

Features

Local speech recognition. Whisper runs in the browser tab (WebGPU when available, otherwise WebAssembly). Pick a model from 40 MB to 250 MB; it downloads once and is cached.

Automatic language detection. Or set the spoken language yourself.

Edit every line. Fix the text of any subtitle before you export.

Translate subtitles into 20 languages. In Chrome this runs on your device for free, using the browser's built-in translation. Timings stay unchanged.

Caption styles. Choose a preset, add word-by-word highlighting, and preview it over your own footage.

Export. Download .srt, .vtt or .ass, or an MP4 with the subtitles burned into the picture, ready for Instagram, TikTok or YouTube Shorts.

Open an existing subtitle file. Import .srt or .vtt to translate it, restyle it or convert between formats, no video needed.

Optional: your own API key. For a larger model or a weak device, send the audio to OpenAI, Groq or Google Gemini with your own key. This is off by default.

Who is it for?

Creators adding captions to short-form video

Anyone subtitling recordings they don't want to upload: lectures, interviews, client calls, family videos

Translators and editors who need a quick SRT/VTT conversion or translation