Loading...
Found 1 Skills
Real-time streaming speech-to-text via Deepgram WebSocket API — sub-300 ms latency, Nova-2 model, speaker diarization, auto-reconnect.