LangSmith 部署

当你准备将 LangChain 智能体部署到生产环境时，LangSmith 提供了一个专为智能体工作负载设计的托管平台。传统的托管平台是为无状态、短生命周期的 Web 应用构建的，而 LangGraph 是专为有状态、长时间运行的智能体构建的，它们需要持久化状态和后台执行。LangSmith 处理基础设施、扩展和运维问题，让你可以直接从代码仓库进行部署。

前提条件

在开始之前，请确保你具备以下条件：

• 一个 GitHub 账号
• 一个 LangSmith 账号（免费注册）

部署你的智能体

1. 在 GitHub 上创建代码仓库

你的应用代码必须存放在 GitHub 代码仓库中才能在 LangSmith 上部署。公开和私有仓库均受支持。对于本快速入门，首先按照本地服务器设置指南确保你的应用与 LangGraph 兼容。然后，将代码推送到仓库。

2. Deploy to LangSmith

1. Navigate to LangSmith Deployment

   Log in to LangSmith. In the left sidebar, select Deployments.

2. Create new deployment

   Click the + New Deployment button. A pane will open where you can fill in the required fields.

3. Link repository

   If you are a first time user or adding a private repository that has not been previously connected, click the Add new account button and follow the instructions to connect your GitHub account.

4. Deploy repository

   Select your application’s repository. Click Submit to deploy. This may take about 15 minutes to complete. You can check the status in the Deployment details view.

3. Test your application in Studio

Once your application is deployed:

1. Select the deployment you just created to view more details.
2. Click the Studio button in the top right corner. Studio will open to display your graph.

4. Get the API URL for your deployment

1. In the Deployment details view in LangGraph, click the API URL to copy it to your clipboard.
2. Click the URL to copy it to the clipboard.

5. Test the API

You can now test the API:

Python

1. Install LangGraph Python:

pip install langgraph-sdk

2. Send a message to the agent:

from langgraph_sdk import get_sync_client # or get_client for async

client = get_sync_client(url="your-deployment-url", api_key="your-langsmith-api-key")

for chunk in client.runs.stream(
    None,    # Threadless run
    "agent", # Name of agent. Defined in langgraph.json.
    input={
        "messages": [{
            "role": "human",
            "content": "What is LangGraph?",
        }],
    },
    stream_mode="updates",
):
    print(f"Receiving new event of type: {chunk.event}...")
    print(chunk.data)
    print("\n\n")

Rest API

curl -s --request POST \
    --url <DEPLOYMENT_URL>/runs/stream \
    --header 'Content-Type: application/json' \
    --header "X-Api-Key: <LANGSMITH API KEY> \
    --data "{
        \"assistant_id\": \"agent\", `# Name of agent. Defined in langgraph.json.`
        \"input\": {
            \"messages\": [
                {
                    \"role\": \"human\",
                    \"content\": \"What is LangGraph?\"
                }
            ]
        },
        \"stream_mode\": \"updates\"
    }"

LangSmith offers additional hosting options, including self-hosted and hybrid. For more information, please see the Platform setup overview.

---

将这些文档连接到 Claude、VSCode 等工具，通过 MCP 获取实时回答。

在 GitHub 上编辑此页面或提交问题。