Hi Product Hunt, I'm the maker of LiveSub. I kept running into the same problem: a stream or meeting in a language I only half follow, and no subtitles. Most tools either only work on one site (YouTube only, Netflix only), or they charge by the minute, because they run speech recognition on their servers. LiveSub does it differently. You share one browser tab, and the speech recognition (Whisper) runs inside your own browser. Translation uses the on-device translator built into Chrome. Your audio is never uploaded, and since there is no per-minute server cost, there is no reason to meter it. That's why it's free with no time limit, and why there is no list of supported sites: if a tab makes sound, it can be subtitled. Being honest about the limits: - Desktop Chrome or Edge only. Phones can't share tab audio - English → Japanese and Japanese → English only, for now - Subtitles appear about 1–2 seconds after speech - Names are the weak point (there's a word list to fix them) - The first run downloads the recognition model (about 206MB for the default English setting); later visits use the cached copy There's a floating always-on-top window so you can watch fullscreen, and you can download the transcript. I'd love to hear where it breaks for you: which sites, which kinds of audio, and whether the tab-sharing step was clear.

LiveSub

Live subtitles for any browser tab, translated on-device

Live subtitles for any browser tab, translated on-device