Pipecat Voice Agent

Smallest AI's STT and TTS are natively integrated into Pipecat, the open-source framework for real-time voice + AI pipelines. This example wires them into a fully working voice agent that runs in your browser via WebRTC. Speak, and the agent responds using Smallest AI's low-latency sophia voice. Start speaking while it's talking — the response stops mid-sentence and the agent picks up your new input immediately. Interruption is handled automatically by Pipecat's SileroVADAnalyzer; no custom logic needed.