LangSmith Deployment: MCP Support For Your LangGraph Agents

Note: LangGraph Platform is now LangSmith Deployment. In this video, we'll show how LangSmith Deployment integrates with MCP (Model Context Protocol). This enables every LangGraph agent to function as an MCP server instantly -- no custom code or infrastructure needed. You'll learn how our built-in MCP support allows developers to deploy and connect LangGraph agents using MCP's streamable HTTP spec, making integration with compatible clients seamless and fast. Overview 00:00 Intro 00:20 MCP Overview 01:02 LangChain & MCP 01:21 Creating a New Deployment 02:37 API URL & MCP Docs 03:39 Configuring Your MCP Client 04:02 Native Tracing Support 04:38 Multi-Agent Support 05:40 Outro Resources - LangSmith Deployment MCP Adapter: https://langchain-ai.github.io/langgraph/concepts/server-mcp/ - MCP Streamable HTTP Spec: https://modelcontextprotocol.io/specification/2025-03-26/basic/transports#streamable-http - LangGraph Documentation: https://langchain-ai.github.io/langgraph/ Whether you're a developer aiming for rapid prototyping or an enterprise seeking scalable AI solutions, LangSmith Deployment's native MCP support offers a streamlined path to leveraging deployed intelligent agents in minutes. #LangGraph #MCP #LangGraphPlatform #LangChain #LLM #AIAgents #MachineLearning #ArtificialIntelligence #DevTools