Pipecat
Open-source framework for real-time voice and multimodal AI agents — pipelines, transports, and tool calling included.
About
Pipecat is the open-source Python framework for building real-time voice and multimodal conversational agents, maintained by the team at Daily. You compose pipelines from modular processors — speech recognition, LLMs, text-to-speech, transports — and Pipecat handles the hard real-time parts: ultra-low-latency audio/video over WebSockets or WebRTC, interruption handling, and orchestration. A single voice assistant or a full multi-agent system (specialists that hand off, fan out in parallel, or coordinate over a shared bus, locally or distributed) is the same programming model.
The ecosystem is unusually complete for the category: official client SDKs for JS, React, React Native, iOS, Android, C++, and ESP32; Pipecat Flows for structured conversations; a CLI that scaffolds projects and deploys to Pipecat Cloud; a real-time debugger (Whisker); and Claude Code skills for scaffolding. At 16,500+ stars and BSD-2-Clause licensed, it is the reference stack teams reach for when a voice agent has to work in production, not just in a demo. The cost is inherent complexity: real-time media pipelines are harder than chat endpoints, and you still bring your own STT/LLM/TTS providers.
What it can do
- Composable real-time pipelines: STT, LLM, TTS, and custom processors
- Ultra-low-latency voice over WebSockets and WebRTC with interruption handling
- Multi-agent systems: handoff, parallel fan-out, sidecar workers, distributed deploys
- Pipecat Flows for structured, state-managed conversations
- Client SDKs: JS, React, React Native, iOS, Android, C++, ESP32
- CLI scaffolding, Pipecat Cloud deploy, Whisker debugger, Claude Code skills
Setup
- Install:
pip install pipecat-ai(oruv tool install "pipecat-ai[cli]"for the CLI). - Run
pipecat init quickstartto scaffold a project (an AI coding assistant can build it with you). - Configure your STT, LLM, and TTS services plus transport (WebSocket or WebRTC).
- Compose processors into a pipeline and run your bot locally.
- Deploy with the CLI to Pipecat Cloud or your own infrastructure.
Prerequisites
Python 3.10+; API keys for STT/LLM/TTS providers (or local models); a transport endpoint for clients.
Use cases
Works with
Trust score — 9/10
Scored 9/10 (verified tier, metrics-v1). Evidence: 16,509 GitHub stars (GitHub API), BSD-2-Clause, pushed Oct 15 2026; the README documents the pipeline model, the SDK matrix, Flows, the CLI, and the debugger; docs corroborate the quickstart and deploy flow. Strengths: the reference open stack for production voice agents, unusually complete ecosystem, Daily-backed. Docked for: (1) real-time media pipelines are inherently harder than chat; (2) you bring your own STT/LLM/TTS providers and their bills. License: BSD-2-Clause.
★ 16,509 GitHub stars
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 8.0/10
- Maintenance · recency, releases, issue load
- 9.0/10
- Openness · license + docs
- 10.0/10
- Manual review · sanity read + red-flag check
- 9.0/10
- · Adoption 8: starScore from 16,509 stars on pipecat-ai/pipecat
- · Maintenance 9: pushed 2026-10-15 (recency 10); frequent releases (10); 398 open issues on 16.5k stars (issues 5)
- · Openness 10: BSD-2-Clause (0.7*10) + official docs verified at docs.pipecat.ai (0.3*10)
- · Manual 9: README documents the pipeline model, 7-platform SDK matrix, Flows, CLI, and Whisker debugger; docked for inherent real-time media complexity and bring-your-own provider costs
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.