Pipecat
Also known as: Pipecat AI, pipecat-ai
Open-source framework for building real-time voice and multimodal AI agents.
Pipecat is an open-source Python framework for building real-time voice and multimodal conversational AI agents. Created by Daily (the real-time audio/video infrastructure company) and maintained with community support, it is deliberately vendor-neutral: developers assemble their own pipeline of speech-to-text, LLM, and text-to-speech services rather than being locked to a single provider.
A Pipecat application is built around a pipeline of processors that stream audio, text, and video frames with ultra-low latency. The framework handles the hard parts of real-time voice (voice-activity detection, turn-taking, interruption handling, and transport) so teams can focus on agent behavior. It supports multiple transports (Daily WebRTC, LiveKit, SmallWebRTC, Twilio/SIP for phone, WebSockets, WhatsApp) and integrates with 100+ AI services across STT (AssemblyAI, Deepgram, Whisper), LLMs (OpenAI, Anthropic, Gemini, Groq, and more), and TTS (Cartesia, ElevenLabs, and others). Client SDKs are available for JavaScript, React, React Native, iOS (Swift), Android (Kotlin), C++, and even ESP32, so agents can connect from web, mobile, and embedded devices.
Beyond single voice bots, Pipecat is multi-agent ready: each pipeline is an agent, and specialists can hand off, fan out in parallel, run as sidecars over a shared bus, or be distributed across processes and machines. Pipecat Flows adds structured conversation paths and state management, and tooling includes a CLI, the Voice UI Kit, and the Whisker debugger. Common applications include voice assistants, customer-support and receptionist bots, AI companions and coaches, and interactive multimodal experiences. The framework is free to self-host; Daily also offers Pipecat Cloud, a managed platform for deploying and scaling Pipecat agents in production.
Vendor details
Canonical URL
https://pipecat.ai
Category
Agent infrastructure
Subcategory
Voice & multimodal agent framework
Funding status
Open-source project created and maintained by Daily (daily.co), a real-time audio/video infrastructure company founded in 2016; developed with community support. Commercial managed hosting offered separately via Pipecat Cloud.
Company status
independent
Use cases & customers
Primary use cases
Target customers
Deployment options
Integrations
Vendor-neutral orchestration across 100+ AI services — STT (AssemblyAI, Deepgram, Whisper, and others), LLMs (OpenAI, Anthropic, Gemini, Groq, and more), and TTS (Cartesia, ElevenLabs, and others). Transports include Daily WebRTC, LiveKit, SmallWebRTC, Twilio/SIP (phone), WebSockets, and WhatsApp. Client SDKs for JavaScript, React, React Native, iOS (Swift), Android (Kotlin), C++, and ESP32. Includes Pipecat Flows (conversation state), a CLI, the Voice UI Kit, and the Whisker debugger.
In practice
Building a voice agent means wrestling with interruptions, turn-taking, and latency before you even get to behavior. Pipecat handles the hard real-time parts, so you focus on what the agent does.
You don't want to be locked to one vendor's speech-to-text or voice. Pipecat is vendor-neutral, letting you assemble your own pipeline across more than 100 STT, LLM, and TTS services.
Your voice agent needs to reach phones, the web, and embedded devices. Pipecat ships client SDKs across web, mobile, and even ESP32, so one agent connects from many surfaces.
Sources & related URLs
Agentic Index coverage score
10.0 / 14 capabilities · 71%
| Integrations & Tool CallingPipecat framework docs, Pipecat Cloud (by Daily) product and deployment pages, and 2026 platform reviews confirm native LLM tool calling with 100+ service integrations, Pipecat Flows orchestration with multi agent handoffs, broad transport and channel coverage (WebRTC, SIP and PSTN telephony, WebSockets, WhatsApp), vendor neutral model choice across any LLM, STT, and TTS, Python plus multi platform client SDKs with MCP support, native OpenTelemetry observability, and self host plus multi region managed deployment. | Full |
|---|---|
| Workflow OrchestrationPipecat framework docs, Pipecat Cloud (by Daily) product and deployment pages, and 2026 platform reviews confirm native LLM tool calling with 100+ service integrations, Pipecat Flows orchestration with multi agent handoffs, broad transport and channel coverage (WebRTC, SIP and PSTN telephony, WebSockets, WhatsApp), vendor neutral model choice across any LLM, STT, and TTS, Python plus multi platform client SDKs with MCP support, native OpenTelemetry observability, and self host plus multi region managed deployment. | Full |
| Knowledge Grounding & RAGKnowledge grounding and long term memory are supported by building retrieval and memory steps into the pipeline or integrating third party services (context aggregators, a memory module, vector stores, and Pipecat Cloud persistent transcript storage) rather than as a managed knowledge base, and prebuilt content is provided as starter examples, the Voice UI Kit, and quickstart bots rather than a large agent marketplace. | Partial |
| Human Oversight & GuardrailsHuman handoff and guardrails are assembled from framework processors and integrations rather than a managed policy layer, security and compliance (HIPAA and GDPR) come through the managed Pipecat Cloud rather than a full identity and governance suite with SOC 2 for Pipecat Cloud still on the roadmap, and testing relies on the Whisker debugger, the CLI, and integrations with simulation and evals tooling rather than a native evaluation product. | Partial |
| Security, Identity & GovernanceHuman handoff and guardrails are assembled from framework processors and integrations rather than a managed policy layer, security and compliance (HIPAA and GDPR) come through the managed Pipecat Cloud rather than a full identity and governance suite with SOC 2 for Pipecat Cloud still on the roadmap, and testing relies on the Whisker debugger, the CLI, and integrations with simulation and evals tooling rather than a native evaluation product. | Partial |
| Observability & AuditabilityPipecat framework docs, Pipecat Cloud (by Daily) product and deployment pages, and 2026 platform reviews confirm native LLM tool calling with 100+ service integrations, Pipecat Flows orchestration with multi agent handoffs, broad transport and channel coverage (WebRTC, SIP and PSTN telephony, WebSockets, WhatsApp), vendor neutral model choice across any LLM, STT, and TTS, Python plus multi platform client SDKs with MCP support, native OpenTelemetry observability, and self host plus multi region managed deployment. | Full |
| Memory & State PersistenceKnowledge grounding and long term memory are supported by building retrieval and memory steps into the pipeline or integrating third party services (context aggregators, a memory module, vector stores, and Pipecat Cloud persistent transcript storage) rather than as a managed knowledge base, and prebuilt content is provided as starter examples, the Voice UI Kit, and quickstart bots rather than a large agent marketplace. | Partial |
| Deployment & Data ResidencyPipecat framework docs, Pipecat Cloud (by Daily) product and deployment pages, and 2026 platform reviews confirm native LLM tool calling with 100+ service integrations, Pipecat Flows orchestration with multi agent handoffs, broad transport and channel coverage (WebRTC, SIP and PSTN telephony, WebSockets, WhatsApp), vendor neutral model choice across any LLM, STT, and TTS, Python plus multi platform client SDKs with MCP support, native OpenTelemetry observability, and self host plus multi region managed deployment. | Full |
| Prebuilt Agents, Templates & PacksKnowledge grounding and long term memory are supported by building retrieval and memory steps into the pipeline or integrating third party services (context aggregators, a memory module, vector stores, and Pipecat Cloud persistent transcript storage) rather than as a managed knowledge base, and prebuilt content is provided as starter examples, the Voice UI Kit, and quickstart bots rather than a large agent marketplace. | Partial |
| Triggers & Channel CoveragePipecat framework docs, Pipecat Cloud (by Daily) product and deployment pages, and 2026 platform reviews confirm native LLM tool calling with 100+ service integrations, Pipecat Flows orchestration with multi agent handoffs, broad transport and channel coverage (WebRTC, SIP and PSTN telephony, WebSockets, WhatsApp), vendor neutral model choice across any LLM, STT, and TTS, Python plus multi platform client SDKs with MCP support, native OpenTelemetry observability, and self host plus multi region managed deployment. | Full |
| Model Flexibility & RoutingPipecat framework docs, Pipecat Cloud (by Daily) product and deployment pages, and 2026 platform reviews confirm native LLM tool calling with 100+ service integrations, Pipecat Flows orchestration with multi agent handoffs, broad transport and channel coverage (WebRTC, SIP and PSTN telephony, WebSockets, WhatsApp), vendor neutral model choice across any LLM, STT, and TTS, Python plus multi platform client SDKs with MCP support, native OpenTelemetry observability, and self host plus multi region managed deployment. | Full |
| APIs, SDKs & MCP ExtensibilityPipecat framework docs, Pipecat Cloud (by Daily) product and deployment pages, and 2026 platform reviews confirm native LLM tool calling with 100+ service integrations, Pipecat Flows orchestration with multi agent handoffs, broad transport and channel coverage (WebRTC, SIP and PSTN telephony, WebSockets, WhatsApp), vendor neutral model choice across any LLM, STT, and TTS, Python plus multi platform client SDKs with MCP support, native OpenTelemetry observability, and self host plus multi region managed deployment. | Full |
| Testing, Debugging & OptimizationHuman handoff and guardrails are assembled from framework processors and integrations rather than a managed policy layer, security and compliance (HIPAA and GDPR) come through the managed Pipecat Cloud rather than a full identity and governance suite with SOC 2 for Pipecat Cloud still on the roadmap, and testing relies on the Whisker debugger, the CLI, and integrations with simulation and evals tooling rather than a native evaluation product. | Partial |
| Browser & Computer UsePipecat targets real time voice and multimodal conversation and does not provide browser or computer use automation. | Unable to verify |
The Agentic Index coverage score grades every vendor Full, Partial or Unable to verify against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded
Recent platform changes
LangChain launched native Python tracing integrations in LangSmith for Pipecat voice agents. The integration captures end-to-end voice pipeline telemetry, including conversation audio, speech-to-text and text-to-speech latency, voice activity detection events, interruptions, and tool calls.
Bears on: Observability / auditability
View sourcePipecat v1.4.0 added realtime_service_mode on LLMContextAggregatorPair, the on_user_turn_message_added event, and RealtimeServiceMetadataFrame, changing how realtime speech-to-speech services write context and expose turn behavior.
Bears on: Workflow orchestration
View sourcePipecat v1.4.0 added the pipecat create project-scaffolding CLI via an optional cli extra, including optional Pipecat Cloud enablement.
Bears on: Deployment / data residency
View sourcePricing
Free (open source) · Cloud from $0.01/agent min
usage
Included quota
The open source framework has no usage limits when self hosted. Pipecat Cloud includes built in Daily WebRTC transport at no added cost, with one human and one bot voice minute free per session, then charges $0.01 per running agent minute. Recording, transcription, and PSTN or SIP telephony add metered charges. Reserved agents and enterprise support are optional add ons.
What is public
Public: the open source framework is free, and Pipecat Cloud's core rate of $0.01 per running agent minute, the reserved instance rate, and telephony add on rates. Not public: enterprise bundled inference pricing and single tenant VPC pricing, which are quoted by sales.
Billing mechanics
The framework is free to self host. Pipecat Cloud is pure usage billing at $0.01 per running agent minute, with optional reserved instances at a lower per minute rate to remove cold starts, plus metered telephony (SIP and PSTN) and a per event call transfer fee. Model, speech to text, and text to speech usage is billed by the underlying providers unless bundled at enterprise scale.
Cost watchouts
Model, speech to text, and text to speech costs are billed by the underlying providers and are usually the largest line item, separate from the $0.01 agent minute rate. Reserved agents that avoid cold starts bill even when idle, telephony and recording are metered, and call transfers carry a per event fee.
Variable cost rationale
Pipecat Cloud bills purely by running agent minutes, and voice workloads run continuously during a call, so cost tracks directly with call volume and duration. Telephony, recording, and transcription add metered charges, and model, speech to text, and text to speech usage is billed by the underlying providers, so total spend at scale is spread across several meters. Self hosting the framework shifts this cost to your own infrastructure and providers.
Additional watchouts
The headline $0.01 per agent minute covers hosting only; model and speech provider costs are separate and typically larger. Reserved agents bill while idle. SOC 2 for Pipecat Cloud is on the roadmap, though the service is HIPAA and GDPR compliant and Daily's underlying WebRTC infrastructure is SOC 2 compliant.
Overage / add-ons
There are no plan caps; usage bills monthly by running agent minutes and metered add ons. Reserved instances bill continuously at a lower rate to keep warm capacity available.
Sales call required
No, self serve available
Free / trial
The Pipecat framework is free and open source under an MIT license and can be self hosted at no cost. Pipecat Cloud uses pay as you go billing and includes free Daily WebRTC transport for development.
Lowest paid plan
Pipecat Cloud pay as you go at $0.01 per running agent minute, above a free and open source self hosted framework.
Commercial notes
A developer first, vendor neutral model: the framework is free and portable, and Pipecat Cloud adds managed scaling and telephony with no lock in, since the same code self hosts. Attractive for teams that want low platform cost and full control over their model and voice stack.
Key ambiguities
Real cost depends heavily on the LLM, speech to text, and text to speech providers a team chooses, which are billed outside the $0.01 agent minute rate, plus telephony and recording. Enterprise bundled inference and VPC pricing are not public.
Missing data
Enterprise bundled inference pricing, single tenant VPC pricing, and exact telephony rates by region are not fully public.
Related vendors
- Acrab — Singapore compute infrastructure company building a full stack…
- AgentOps — Agent observability and reliability platform with broad model and…
- Agno — High-performance agent runtime and framework (formerly Phidata) with…
- AIsa — Unified resource and payment gateway for AI agents that lets them…
- AlphaBitCore — AI control plane that governs how models, agents, tools, and…
- Anchor Browser — Cloud hosted browser infrastructure that lets AI agents operate real…
Alternatives to Pipecat
The closest documented capability profiles to Pipecat among agent infrastructure platforms tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.
- Bernstein10.5 / 14Fuller documented coverage on Security, Identity & Governance and Testing, Debugging & Optimization
- Paragon11.5 / 14Fuller documented coverage on Knowledge Grounding & RAG and Security, Identity & Governance
- Tray.ai11.5 / 14Fuller documented coverage on Human Oversight & Guardrails and Security, Identity & Governance
- Vapi10.5 / 14Fuller documented coverage on Security, Identity & Governance and Testing, Debugging & OptimizationPipecat vs Vapi →
- Waniwani9.5 / 14Fuller documented coverage on Knowledge Grounding & RAG
- xpander.ai10.5 / 14Adds documented Browser & Computer Use
Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded