Subtitles for your video — nothing uploads
Drop a video or audio file and get an SRT / VTT / TXT file back. Everything runs on your device. No sign-up, no upload, no limit.
Advanced options
This tool is built and verified for English. Other languages run on the same small model and are noticeably less accurate — expect to correct a lot of the output. Chinese comes out in Traditional characters.
Your subtitles
Click any line to jump the player. Edit text directly — changes apply to the download.
How it works
- Drop a file. It stays on your device — nothing is uploaded, ever.
- First run only: the speech model (~75 MB) downloads once and is cached by your browser.
- Transcribe. Roughly 0.5–1× realtime on a desktop: under a minute runs about realtime, a 2-minute clip takes about 3 minutes.
- Export. SRT (Premiere / Resolve / YouTube), VTT (HTML5) or plain text.
After the first download the model is cached, so you can use the tool completely offline. Open the Network tab in DevTools if you want to verify: no request ever carries your audio.
FAQ
Is my file uploaded anywhere?
No. Transcription runs on your machine using WebAssembly. You can disconnect from the internet after the model is cached and it keeps working.
Why does the first use download 75 MB?
That is the speech recognition model. It downloads once and is then cached by your browser, so later visits need no download at all.
How long does it take?
Under a minute runs at about realtime. A 2 minute clip takes roughly 3 minutes, and a 10 minute clip around 20 minutes — the longer the file, the lower the multiple. The tool shows an estimate before it starts, and it becomes more accurate after your first run. For long recordings we suggest working in chunks of a few minutes.
Which formats can I export?
SRT, WebVTT and plain text. SRT and VTT keep the timings; TXT gives you the transcript only.
Can I process several files at once?
Yes. Drop two or more files at once and they go into a batch queue. The files are transcribed one after another, and each one gets its own SRT, VTT and TXT download button the moment it is finished — you do not have to wait for the whole batch. The speech model is downloaded once for the entire batch, so the rest of the files start straight away. Keep the tab open until the queue is done.
Do I need to sign up?
No. There is no account, no daily quota and no length limit.