Video Subtitles
Subtitle a video of any length without uploading it — the video never leaves your device. We find the cast list on Wikipedia and Wikidata, which helps the speech model spell the character names. See real use cases →
Private by design: your video file never leaves this device. The browser extracts the audio locally and uploads only small compressed audio pieces for transcription. Works for videos of any length — no account needed.
The AI subtitle generator you pay with Bitcoin
Every other subtitle generator wants an account, a subscription, and your whole video uploaded to their cloud. This one takes none of those: the browser extracts the audio from your video locally, sends only small compressed audio pieces for AI transcription, and stitches the results into one SRT/VTT file — for a video of any length, including full movies. You pay 10 sats per minute of audio over the Bitcoin Lightning Network, with no account and no subscription.
Character names from the cast list
A speech model gets names wrong more often than any other word. When you choose a file, we look its title up on Wikipedia and Wikidata and offer the film's character names and places before you pay. Remove a name, add your own, or search for another film. For this lookup, only the title and the year from the file name leave your browser.
Private by architecture, not by promise
Because subtitles only need the audio (about 1% of a video file), the video itself never leaves your device. Transcripts are deleted from our servers by the hourly cleanup, 2 hours after the job finishes. Files with multiple audio tracks — original language, dubs, commentary — show a track picker before you pay, and an interrupted job resumes where it stopped without paying twice.
Also see Speech to Text for plain audio transcription with word-level timestamps.