Build AI Workflows
Visually. No Code.
Drag, connect, and run AI agent workflows from a visual canvas. Works with any major LLM — OpenAI, Anthropic, Gemini, and more. Parallel execution, live streaming, and a public API — no code required.
No credit card required.
run #a3f9 · customer-support
The idea
The agent graph, made tangible.
Make the graph visible
Every agent is just LLM calls, tool calls, and routing logic. Pipecat puts that graph on a canvas where you can see, edit, and debug every edge in real time.
Real parallelism
Nodes with no shared dependencies run concurrently. A three-branch graph finishes in the time of its slowest branch — not the sum of all three.
One click to a public API
Enable a workflow, get a POST /flows/{slug}/invoke endpoint with an API key. Sync, async, or SSE stream. Call it from anywhere in one line of curl.
Capabilities
Everything you need to build powerful workflows.
Visual DAG Builder
Drag and drop nodes onto an infinite canvas. Connect them into directed acyclic graphs that execute in topological order.
Real-time Streaming
WebSocket events stream live node status directly to the canvas as each step completes.
Schedule Trigger
Run any workflow on a cron schedule — every 15 minutes, daily, weekdays only, or a fully custom expression.
asyncio.gather — 3x faster
Parallel Execution
Nodes without dependencies run concurrently, cutting total latency to the slowest branch — not the sum.
POST https://api.your-tool.com/run
{
"query": "{input}"
"limit": 10
}
// result returned to the agent
Custom HTTP Tools
Register any HTTP endpoint as a tool. The agent calls it automatically mid-run — no extra code.
Form Fields
Form Trigger
Build a shareable form with text, email, and select fields. Each submission triggers the workflow with the captured values.
$ curl -X POST
https://pipecat.dev/flows/support/invoke
-H "X-Api-Key: sk-…"
-d '{"input": "Where is my order?"}'
// → { result: "Your order ships…" }
Public Invoke API
Deploy any workflow and get a public REST endpoint with API-key auth. Invoke sync, async, or via SSE stream from any system.
Integrations
Works with the tools you already use.
Connect Gmail, Google Docs, HubSpot, and Calendly as LLM tools — the agent calls them automatically mid-run, no plumbing required.
Need something else? Register any HTTP endpoint as a tool — works with anything that has an API.
Pricing
Simple, predictable pricing.
Start free. Scale when you're ready.
Hobby
For side projects and exploration.
- 100 workflow runs / month
- 3 active workflows
- 5 custom tools
- Real-time WebSocket streaming
- Community support
Pro
~$19 USD
For teams shipping production AI.
- 10,000 workflow runs / month
- Unlimited workflows & tools
- Public Invoke API + API keys
- Async runs + webhook callbacks
- SSE streaming endpoint
- Priority support
Enterprise
Unlimited scale, dedicated infrastructure.
- Unlimited runs
- Dedicated compute
- SSO / SAML
- Audit logs
- Custom model endpoints
- SLA guarantee
- Dedicated Slack channel
Questions
Frequently asked questions.
What is a visual AI agent workflow?+
A visual AI agent workflow is a directed graph of LLM calls, tool calls, and routing logic that you build by dragging and connecting nodes on a canvas. Pipecat runs the graph for you with real parallelism, streams execution events live, and exposes each workflow as a public REST API.
Does Pipecat replace LangChain?+
Yes for most use cases. Pipecat covers the orchestration layer — LLM invocations, tool calling, branching, parallel execution, and streaming — without LangChain's abstraction overhead. The runtime is a few hundred lines of Python on top of the major LLM SDKs.
Can I expose any workflow as a REST API?+
Yes. Enable a workflow and Pipecat gives you a POST /flows/{slug}/invoke endpoint with an API key. You can invoke it synchronously, asynchronously with a webhook callback, or as an SSE stream from any system.
Does it support tool use with Claude, GPT-4, and Gemini?+
Yes. Pipecat works with any major LLM — Anthropic Claude, OpenAI GPT-4 / GPT-4o, Google Gemini, and others. Register any HTTP endpoint as a tool and the agent calls it automatically mid-run.
How is Pipecat different from Zapier or n8n?+
Zapier and n8n are general workflow automation tools. Pipecat is purpose-built for AI agents — it understands LLM nodes, tool calls, the agentic loop, and streaming execution. Parallel nodes run with asyncio.gather rather than sequentially.
Is there a free plan?+
Yes. The Free plan includes 100 workflow runs per month, 3 active workflows, 5 custom tools, real-time WebSocket streaming, and community support. No credit card required.
How do I deploy a workflow built in Pipecat?+
Deployment is one click. Click Enable on any workflow to receive a public REST endpoint with API-key auth. There is no infrastructure to manage — Pipecat hosts the runtime.
Running an online store?
Pipecat also ships a drop-in AI shopping assistant — semantic product search, add to cart, and in-chat checkout.
Ship an AI pipeline in 10 minutes.
No infrastructure to manage. No LangChain to fight. Just nodes, edges, and results.
Start for free