Whisper Live
Open-source neural STT based on Whisper. Strong accuracy, multilingual.
Whisper Live is an open-source neural STT engine based on OpenAI's Whisper architecture. Strong accuracy, multilingual, runs locally.
Setup
Add the service
Manage Services → + Add Services → Whisper Live → Add. Voxta installs the Python runtime and model weights automatically on first use.
Pick a model size
In the config, choose your Whisper model size (tiny / base / small / medium / large). Bigger models = better accuracy + slower + more VRAM. small or medium is a good starting balance.