Twinsubs
Audio to subtitles: make an SRT in your browser
Drop in an audio or video file and get an SRT subtitle file a few minutes later. Recognition runs in your own browser, so the file never leaves your computer, and there is nothing to sign up for.
- One file at a time, up to 15 minutes
- Original-language subtitles only, no translation
- The first run downloads about 150 MB of model; your browser remembers it after that
This small model is noticeably less accurate in languages other than English. For better accuracy, use the Mac app.
Drop an audio or video file here
or
Subtitle preview
Only the first 200 cues are shown; the downloaded file is complete.
How it works
- Choose or drop one audio or video file (MP3, M4A, WAV, MP4, MOV and so on).
- Wait for recognition to finish; the progress bar keeps moving. The first run downloads the model first.
- Look over the preview, click Download SRT, and import it into CapCut, Final Cut Pro, or your player.
FAQ
Is my file uploaded?
No. Recognition runs in your browser, and the audio and video stay on your computer. The page only downloads the model and its scripts.
Why is the first run slow?
It has to download about 150 MB of speech model first (a little more on some computers). Your browser remembers it, so the next run starts right away.
Which browsers work?
Recent Chrome and Edge are fastest because they use your graphics card. Other browsers fall back to the processor, which is slower, and a 15-minute file can take a long time. Phone browsers are not recommended.
Why only 15 minutes?
This page uses a small model and is limited by browser memory. For longer files, batches, and bilingual subtitles, use the Mac app.
How accurate is it?
Clear speech comes out well; heavy accents, loud music, and people talking over each other cause mistakes. Check the result against the audio before you import it.