LangGraph
LangGraph is a low-level framework for orchestrating long-running, stateful AI agents. Workflows are modelled as a graph – with durable execution, human-in-the-loop and memory.
- Project
- LangGraph
- Publisher
- LangChain
- Repository
- github.com/langchain-ai/langgraph
- Documentation
- docs.langchain.com/oss/python/langgraph/overview
- Languages
- Python, TypeScript
- Category
- AI agent frameworks
What is LangGraph?
LangGraph is an open-source framework by LangChain Inc. for agents and workflows that run for long periods and keep state. A workflow is described as a graph of nodes and edges, so branches, loops and subgraphs can be defined explicitly. LangGraph is available for Python and JavaScript/TypeScript and can be used without LangChain.
Core capabilities
- Durable execution
- Agents survive failures and interruptions and resume exactly where they left off.
- Human-in-the-loop
- An agent’s state can be inspected and modified at any point before it continues.
- Memory
- Short-term memory for the current task and long-term memory across sessions.
- Observability and deployment
- With LangSmith, workflows can be traced, evaluated and operated.
When LangGraph is the right choice
LangGraph fits when you want to define exactly which steps an agent takes, where a person intervenes and what happens on failure. For quick agents with standard behaviour, LangChain is simpler; if you want planning, sub-agents and a filesystem out of the box, use DeepAgents, which is built on LangGraph.
Using LangGraph effectively in an AI-first business model
An AI-first operating model defines which actions an agent prepares, recommends, executes with approval or performs autonomously within defined limits. LangGraph is well suited to this because such approval points can be modelled as interrupts in the graph.
The gradation itself remains a business decision. It depends on how large the impact of an action is, whether it can be reversed and who is accountable for it. These questions belong at the start – only then is it worth building the graph.
From tool to outcome
In AI-first coaching you clarify which outcomes AI agents should be accountable for in your go-to-market and within which framework. REVAGO then designs, builds and runs your system. You pay for results – not for activity.
Frequently asked questions
What is LangGraph?
LangGraph is an open-source, low-level framework for orchestrating long-running, stateful AI agents. Workflows are modelled as a graph; durable execution, human-in-the-loop and short- and long-term memory are built in. LangGraph is made by LangChain Inc. and is available for Python and JavaScript/TypeScript.
Do I need LangChain to use LangGraph?
No. LangGraph can be used on its own. It integrates, however, with LangChain components, with DeepAgents as a higher-level layer and with LangSmith for tracing, evaluation and deployment.
How does LangGraph support human-in-the-loop?
LangGraph can pause a workflow at defined points so that a person can inspect, modify or approve the state. Which actions require approval should be defined in the operating model beforehand.
Further reading in the wiki
A neutral overview, not a recommendation of individual providers. REVAGO is not affiliated with the listed projects. Descriptions are based on the projects’ public information; features and licences may change.