search

KoljaB / RealtimeSTT

A robust, efficient, low-latency speech-to-text library with advanced voice activity detection, wake word activation and instant transcription.
MIT License
2.09k stars 190 forks source link

Client side JavaScript version #145

Open ymerkos opened 2 weeks ago

ymerkos commented 2 weeks ago

B"H Fantastic library

Is there any way to get this working fast on the browser using wasm or something?

KoljaB commented 2 weeks ago

Not really. There are some implementations for whisper running fully in the browser. Yet RealtimeSTT also uses two VAD model and two wakeword detection models. These all don't work solely in the browser.