All Tools
M
OtherFreeOpen Source
MOONSHINE
On-device streaming speech-to-text and text-to-speech toolkit
MIT
ABOUT
Cloud ASR and TTS add latency, cost, and privacy risk for live voice agents. Moonshine runs streaming speech models on-device so you can transcribe while the user is still talking and synthesize replies without an API key.
INSTALL
pip install moonshine-voiceINTEGRATION GUIDE
1. Stream microphone audio into on-device English or multilingual speech-to-text
2. Build a local voice agent that transcribes and speaks without cloud APIs
3. Ship a tiny 1MB speech model on Raspberry Pi or mobile
4. Add JavaScript or WASM speech I/O to a browser voice interface
5. Swap Whisper-class accuracy for lower-latency streaming recognition
TAGS
speech-to-texttext-to-speechvoice-agentson-devicestreamingpythonasrtts