LangGraph
#12langchain · seit 22. Januar 2024 (Erstveröffentlichung); Version 1.0 erschien am 29. Oktober 2025 · 17× · zuletzt 30. Juni 2026
27
Momentum
LangGraph is an open-source agent orchestration framework developed by LangChain Inc. It is a low-level runtime for building, managing, and deploying long-running, stateful agents and multi-agent systems using a graph model of nodes and edges. Core capabilities include durable execution, streaming, human-in-the-loop checkpoints, checkpointing/persistence, and short-/long-term memory, and it is used in production by companies such as Klarna, Uber, and J.P. Morgan. A commercial "LangGraph Platform" additionally provides managed deployment and debugging (LangGraph Studio) with usage-based pricing.
Historique du momentum
04.04.03.07.
Fonctionnalités
| Autonomy Level | Configurable: from fully autonomous agent loops to human-in-the-loop approval points (interrupts) for critical actions |
| Channels | Programmatic access via Python/JS SDK, LangGraph Server/API, LangGraph Studio (visual interface), deployment via LangGraph Platform |
| License | MIT license, open source, free to use (core library) |
| Modalities | Text/message-based (LLM prompts, tool calls); UI update message types available for streaming interfaces |
| Platform | Python and JavaScript/TypeScript library (langgraph / @langchain/langgraph); optional LangGraph Platform (cloud/hybrid/self-hosted) for deployment |
| Price | Open-source library free (MIT); Platform 'Plus': from $39/month/seat (LangSmith Plus) + $0.001 per node execution; Enterprise: custom pricing |
| Release Date | January 22, 2024 (launch); v1.0 on October 29, 2025 |
| Tool/MCP Integrations | Supports MCP via langchain-mcp-adapters; 600+ model/tool integrations in the LangChain ecosystem; compatible with any tool/model provider |
