Repolode
Submit
pipecat-ai

pipecat-ai/pipecat

Open-source Python framework for building real-time voice and multimodal AI agents, with pluggable speech, LLM, and transport services.

View on GitHubWebsite
Stars
16,606
Forks
2,904
Open issues
398
Last commit
14 hours ago

What is it?

Pipecat is a Python framework for building real-time conversational agents that handle voice, video, and other multimodal input. Developers assemble an agent as a pipeline of modular processors covering speech-to-text, LLM calls, text-to-speech, and transport, so each stage can be swapped without rewriting the rest.

The project includes a broad set of integrations: many speech recognition, language model, and text-to-speech providers, plus speech-to-speech services such as OpenAI Realtime and Gemini Multimodal Live. Transports include WebRTC through Daily, LiveKit, and Vonage, as well as WebSocket and FastAPI WebSocket connections, and telephony serializers for providers like Twilio and Telnyx. Support for multi-agent setups, where pipelines hand off to each other or run in parallel, is also described.

A CLI (`pipecat init`) scaffolds new projects, and client SDKs cover JavaScript, React, iOS, Android, and C++. The framework is distributed as the `pipecat-ai` package under the BSD-2-Clause license.

Key features

  • 01Composable pipelines of modular processors for speech, LLM, and TTS stages
  • 02Wide range of speech, LLM, and text-to-speech provider integrations
  • 03Speech-to-speech support through OpenAI Realtime and Gemini Multimodal Live
  • 04WebRTC and WebSocket transports, plus telephony serializers for Twilio and others
  • 05Multi-agent handoff and parallel pipelines
  • 06CLI for scaffolding projects and client SDKs for web and mobile

Use cases

  • Building a voice assistant that answers calls or speaks with users in real time
  • Creating video or multimodal bots that combine speech with vision models
  • Developing structured business agents such as customer intake or support flows
  • Coordinating several specialist agents that hand off conversations

Quick start

  1. 1

    Install the Pipecat CLI

    uv tool install "pipecat-ai[cli]"
  2. 2

    Create a project with the quickstart template

    pipecat init quickstart

Topics

#voice-agent#real-time#llm#speech-to-text#text-to-speech#webrtc#python#multimodal#ai#chatbot-framework

FAQ

What is pipecat?
Open-source Python framework for building real-time voice and multimodal AI agents, with pluggable speech, LLM, and transport services. Pipecat is a Python framework for building real-time conversational agents that handle voice, video, and other multimodal input. Developers assemble an agent as a pipeline of modular processors covering speech-to-text, LLM calls, text-to-speech, and transport, so each stage can be swapped without rewriting the rest.
Is pipecat free to use? What license does it have?
pipecat is open source under the BSD-2-Clause license. Check the license terms before using it in commercial products.
Is pipecat still maintained?
Actively maintained. The last commit to pipecat was on Oct 19, 2026.
What can I use pipecat for?
Building a voice assistant that answers calls or speaks with users in real time Creating video or multimodal bots that combine speech with vision models Developing structured business agents such as customer intake or support flows Coordinating several specialist agents that hand off conversations
How do I install pipecat?
Install the Pipecat CLI: uv tool install "pipecat-ai[cli]" Create a project with the quickstart template: pipecat init quickstart
Where can I get pipecat?
The source code and installation instructions are on GitHub: https://github.com/pipecat-ai/pipecat

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects