From the archive
LangGraph Platform reaches general availability
LangChain adds managed deployment and state storage for agents that run over time and pause for human input.
LangChain announced general availability of LangGraph Platform, its deployment and management service for long-running AI agents. The offering builds on the beta introduced the previous June.
Keep work running between interactions
An agent may need several model calls, outside tools, or a person's decision before it finishes. LangGraph Platform provides storage for the state of that work, conversation history, and memory, alongside scaling for changing request volumes.
Its Studio environment helps inspect the steps of a workflow and revisit a point where execution needs attention. The management interface also keeps track of deployed agents and their versions.
Choose the deployment arrangement
The release offers managed cloud, hybrid, and self-hosted deployment options. That gives teams a choice between a hosted service and retaining more of the infrastructure in their own environment. The LangChain profile covers its development framework, workflow tooling, and deployment services.