What is this?
Pipecat is an open-source Python framework for building real-time voice and multimodal conversational agents — from a single voice bot to a distributed multi-agent system — maintained by Daily and the community.
Pipecat is an open-source Python framework for building real-time voice and multimodal conversational agents — from a single voice bot to a distributed multi-agent system — maintained by Daily and the community.So it's a framework I write code against, not a ready-made app I install?It's a code-first Python framework: you compose agent pipelines from modular components (speech, AI services, transports) and run them server-side.
It's a code-first Python framework: you compose agent pipelines from modular components (speech, AI services, transports) and run them server-side.How do I get my first agent running?Run pipecat init quickstart to get a working agent, or follow the quickstart guide linked in the README.
Run pipecat init quickstart to get a working agent, or follow the quickstart guide linked in the README.Which speech and AI services can I plug into it?It ships adapters for dozens of services: speech-to-text (Deepgram, AssemblyAI, Whisper), LLMs (OpenAI, Anthropic, Gemini, Ollama), text-to-speech (ElevenLabs, Cartesia, Piper) and speech-to-speech (OpenAI Realtime, Gemini Multimodal Live).
It ships adapters for dozens of services: speech-to-text (Deepgram, AssemblyAI, Whisper), LLMs (OpenAI, Anthropic, Gemini, Ollama), text-to-speech (ElevenLabs, Cartesia, Piper) and speech-to-speech (OpenAI Realtime, Gemini Multimodal Live).Is this for developers writing code, or is there a no-code editor?Developers: you build agents by writing Python pipelines; the README mentions no visual or no-code builder.
Developers: you build agents by writing Python pipelines; the README mentions no visual or no-code builder.How does my app or device connect to a Pipecat agent?Clients connect over real-time transports like WebSockets or WebRTC with ultra-low latency, and official client SDKs let you connect from any platform.
Clients connect over real-time transports like WebSockets or WebRTC with ultra-low latency, and official client SDKs let you connect from any platform.What can I actually build with it?The README highlights voice assistants, multi-agent systems, AI companions like coaches or meeting assistants, multimodal interfaces with voice and video, interactive storytelling with generative media, business agents for support and intake, and structured dialog systems.
The README highlights voice assistants, multi-agent systems, AI companions like coaches or meeting assistants, multimodal interfaces with voice and video, interactive storytelling with generative media, business agents for support and intake, and structured dialog systems.Can it run multiple agents that work together?Yes. Each pipeline is an agent, and Pipecat supports specialists that hand off to each other, fan out in parallel, run as sidecar workers, or coordinate over a shared bus — locally or distributed across processes and machines.
Yes. Each pipeline is an agent, and Pipecat supports specialists that hand off to each other, fan out in parallel, run as sidecar workers, or coordinate over a shared bus — locally or distributed across processes and machines.Do I have to use paid cloud APIs, or can it run local models?Local options are among the supported adapters: Ollama for LLMs, Piper and Kokoro for text-to-speech, and Whisper or FunASR for speech-to-text.
Local options are among the supported adapters: Ollama for LLMs, Piper and Kokoro for text-to-speech, and Whisper or FunASR for speech-to-text.Is it fast enough for natural back-and-forth conversation?Pipecat is built for real-time use: it targets ultra-low-latency interaction and supports transports such as WebSockets and WebRTC.
Pipecat is built for real-time use: it targets ultra-low-latency interaction and supports transports such as WebSockets and WebRTC.What are similar projects?agents, jargo, voltagent, ten-framework, ChatTTS, Omni-Interaction-Agent, agenticSeek, hive
agents, jargo, voltagent, ten-framework, ChatTTS, Omni-Interaction-Agent, agenticSeek, hiveWhich topics is it in?chatbot-framework, chatbots, multimodal-ai, ai-companion, conversational-ai, voice-assistant, webrtc, voice, speech-to-text, tts, agent-framework, multimodal
chatbot-framework, chatbots, multimodal-ai, ai-companion, conversational-ai, voice-assistant, webrtc, voice, speech-to-text, tts, agent-framework, multimodal