orukPipecat 1.8.1 · WebRTC

Live words.
Vocal context alongside them.

Send a public sample through the browser and Pipecat to Oruk’s realtime API.
Real speech inference. No microphone access, language model, or synthetic reply.

Ready to connect. The API key stays on the server.

Transcript

Your sample will appear here.

Phrase estimate

Waiting for audio

Scores describe estimated vocal expression. They are not verified feelings or calibrated probabilities.

Input: Oruk’s public quickstart recording · API documentation
Model: oruk-realtime · September 11, 2026 · Development adapter

Session events