Pipecat AI

Pipecat AI is the open-source Python framework for building ultra-low-latency (500-800ms) voice and multimodal conversational agents.

Pipecat is your go-to open-source Python framework for real-time conversational AI. It uses a modular, pluggable pipeline architecture (orchestrating STT, LLMs, and TTS) to build voice assistants, AI companions, and complex business agents. The framework ensures ultra-low latency: typical voice interactions complete in 500-800ms, making conversations feel natural and responsive. Developers integrate vendor-neutral AI services (like OpenAI or Eleven Labs) and deploy across platforms using client SDKs (JavaScript, Swift, C++), focusing on unique agent logic, not orchestration complexity.

https://pipecat.ai