Docs by LangChain home page
Search...
⌘K
LangGraph Platform
Platform for building and deploying AI agents
Overview
Deployment options
Cloud SaaS
Hybrid
Self-hosted
Guides for deployment
Deploy to cloud SaaS
Deploy hybrid
Deploy self-hosted full platform
Deploy self-hosted data plane only
Interact with the deployment using RemoteGraph
Configure your application for deployment
Set up a LangGraph application with requirements.txt
Set up a LangGraph application with pyproject.toml
Set up a LangGraph.js application
Customize the Dockerfile
Rebuild graph at runtime
LangGraph CLI
LangGraph SDK
Egress for metrics and metadata
Forum
Docs by LangChain home page
Search...
⌘K
Ask AI
Forum
Search...
Navigation
Overview
Hybrid
Get Started
Build
Deploy
Manage
Reference
Get Started
Build
Deploy
Manage
Reference
Overview
Hybrid
Copy page
Copy page
Important
The Hybrid deployment option requires an
Enterprise
plan.
Requirements
You use
langgraph-cli
and/or
LangGraph Studio
app to test graph locally.
You use
langgraph build
command to build image.
Hybrid
The
Hybrid
deployment option lets you manage the
data plane
in your own cloud, while we handle the
control plane
in ours. When using the Hybrid version, you authenticate with a
LangSmith
API key.
Control plane
Data plane
What is it?
Control plane UI for creating deployments and revisions
Control plane APIs for creating deployments and revisions
Data plane “listener” for reconciling deployments with control plane state
LangGraph Servers
Postgres, Redis, etc
Where is it hosted?
LangChain’s cloud
Your cloud
Who provisions and manages it?
LangChain
You
For information on how to deploy a
LangGraph Server
to Hybrid, see
Deploy to Hybrid
Architecture
Compute Platforms
Kubernetes
: The Hybrid deployment option supports deploying data plane infrastructure to any Kubernetes cluster.
Amazon ECS
: Coming soon!
If you would like to deploy to Kubernetes, you can follow the
Hybrid deployment guide
.
Previous
Self-hosted
Next
On this page
Requirements
Hybrid
Architecture
Compute Platforms
Assistant
Responses are generated using AI and may contain mistakes.