October 22, 2025
Pipecat 0.0.91 release just out ...
I'm particularly excited about trickle ICE in the serverless, zero-dependencies SmallWebRTCTransport (because, you know, WebRTC), and support for the AWS Nova Sonic speech-to-speech model in the universal LLMContext class.
The last couple of weeks have been busy. Additions to Pipecat that I have barely had time to do more than look at briefly include:
- Deepgram Flux transcription model launch (free in Pipecat Cloud for all of October)
- Krisp VIVA
- Gemini Nano Banana (image generation!)
- Gemini Live API update
- Gemini Live API on Vertex support
- More ESP32 hardware project sample code
- Hume TTS
- HeyGen transport
Changelog[1]
Getting started[2]