Skip to main content
Pipecat Cloud is the platform for running Pipecat agents in production. You bring an agent image; it handles deploys, session starts, scaling, and observability. It comes in two deployment models that share one control plane and differ in where your agents run.

​
Two ways to run

Pipecat Cloud

Multi-tenant and fully managed. Your agents run in Daily-hosted regions and Daily operates everything. Sign up and deploy in minutes.

Pipecat Enterprise

Single-tenant and hybrid. Your agents run in a self-hosted region — a Kubernetes cluster in your VPC.
The Pipecat Enterprise overview has the full comparison, and Architecture explains how the control plane and data plane divide the work.

​
Start on Cloud, grow into Enterprise

Pipecat Cloud and Pipecat Enterprise are one platform. Start on Cloud and move agents to a self-hosted region when data residency, network placement, or scale calls for it. Your image, deploy workflow, and session code carry over unchanged, and you can run agents in both. What stays the same:
  • Agent images: The same image format, pulled from any container registry. See Agent Images.
  • Deploys: pipecat cloud deploy, the dashboard, or pcc-deploy.toml, with a region chosen per agent. See Deployments.
  • Sessions: The same REST API and Python SDK on any transport — Daily WebRTC, WebSocket, or telephony.
  • Scaling: min_agents and max_agents, scaling from zero with session demand. See Scaling.
  • Secrets: Agents reference secret sets by name. In a self-hosted region, the values live in your cluster. See Secrets.
  • Observability: Sessions, per-session CPU and memory metrics, and agent logs in the dashboard, CLI, and API. In a self-hosted region, logs are stored in your own log store, which is optional.
What changes is where things live: secret values move into your cluster, you set agent sizing yourself, and WebSocket and telephony traffic connects to your region’s own endpoint. The Pipecat Enterprise overview lists every difference.

​
How it works

1

Build your agent

Write a Pipecat pipeline as you normally would, using any supported services.
2

Install the CLI

Install the Pipecat CLI with the Pipecat Cloud plugin:
3

Deploy

Build and deploy your agent image to a region — Daily-hosted or your own.
4

Start sessions

Use the REST API or Python SDK to start agent sessions on demand.

​
Pricing

Pipecat Cloud uses usage-based pricing. See the pricing page for current rates and a calculator to estimate your total cost per minute, including third-party AI services. Pipecat Enterprise is billed under your enterprise agreement; agent-minutes in self-hosted regions are not metered by agent profile. Contact sales for details.

​
Next steps

Create your account

Get your account set up and ready to deploy your first agent.

Deploy your first agent

Deploy the quickstart agent to Pipecat Cloud.

Pipecat Enterprise

What you operate in a self-hosted region, and how to get started.