Langgraph Platform Deployment Tabs

<Callout type="info">
  You will need a [LangSmith account](https://smith.langchain.com) to use this method.
</Callout>

<Callout type="warning">
  **Python agents only**: If you encounter a "No checkpointer set" error, prefix your command with `LANGGRAPH_API=true`:

  ```bash
  LANGGRAPH_API=true langgraph dev --host localhost --port 8000
  ```

  This tells your agent to run without a custom checkpointer, relying on LangGraph's built-in state management instead.
</Callout>

```bash
# For Python 3.11 or above
langgraph dev --host localhost --port 8000
```

```bash
# For TypeScript with Node 18 or above
npx @langchain/langgraph-cli dev --host localhost --port 8000
```

After starting the LangGraph server, the deployment URL will be `http://localhost:8000`.

<Accordions className="mb-4">
    <Accordion title="Having trouble?">
        If you cannot run the `langgraph dev` command, or encounter any issues with it, try this one instead:
        `uvx --refresh --from "langgraph-cli[inmem]" --with-editable . --python 3.12 langgraph dev`
    </Accordion>
    <Accordion title='"No checkpointer set" error with langgraph dev'>
      Python agents automatically detect their runtime environment, but sometimes need explicit guidance:

      - **With `langgraph dev`**: Set `LANGGRAPH_API=true` to disable custom checkpointer
      - **With local FastAPI**: Don't set this flag; uses MemorySaver automatically
      - **JavaScript agents**: Not affected by this issue
    </Accordion>
</Accordions>

1. Add a `.env` file to contain your LangSmith API key.

```plaintext title=".env"
LANGSMITH_API_KEY=your_langsmith_api_key
```

2. Next, use CopilotKit's FastAPI integration to serve your LangChain agent.

```python title="server.py"
import os
from fastapi import FastAPI
import uvicorn
from copilotkit import LangGraphAGUIAgent # [!code highlight]
from ag_ui_langgraph import add_langgraph_fastapi_endpoint # [!code highlight]
from sample_agent.agent import graph # the coagents-starter path, replace this if its different # [!code highlight]

from dotenv import load_dotenv
load_dotenv()

app = FastAPI()
# [!code highlight:9]
add_langgraph_fastapi_endpoint(
  app=app,
  agent=LangGraphAGUIAgent(
    name="sample_agent", # the name of your agent defined in langgraph.json
    description="Describe your agent here, will be used for multi-agent orchestration",
    graph=graph, # the graph object from your langgraph import
  ),
  path="/your-agent-path", # the endpoint you'd like to serve your agent on
)

# add new route for health check
@app.get("/health")
def health():
    """Health check."""
    return {"status": "ok"}

def main():
    """Run the uvicorn server."""
    port = int(os.getenv("PORT", "8000"))
    uvicorn.run(
        "sample_agent.demo:app", # the path to your FastAPI file, replace this if its different
        host="0.0.0.0",
        port=port,
        reload=True,
    )

```

Come back with that URL and a LangSmith API key before proceeding.

<Callout type="info">
  A successful deployment will yield an API URL (often referred to here as "deployment URL").
  In LangGraph Platform, it will look like this: `https://{project-identifiers}.langgraph.app`.
</Callout>