Open sourcePython · ≈45k stars · MIT
LangGraph
Low-level orchestration for controllable, stateful agents.
LangGraph models agents as graphs of nodes with checkpointed state, cycles, human-in-the-loop interrupts and durable execution. It has become the reference architecture for production agent orchestration.
Pros & cons
Each point cites its sources and says what kind of evidence backs it. A second reviewer tried to refute every claim before it was kept.
Pros
- User reportsUsers running it in production cite checkpointing, pausing mid-graph for human input and time-travel over state as the main reasons to use it.[1][2][3]
- User reportsUsers find the explicit graph/state-machine model a useful way to control agent flow, even when they dislike the wider LangChain abstractions.[1][2]
- User reportsThe Postgres checkpointer lets stateless backends resume sessions without session affinity.[1]
- MetadataMIT licensed, actively released (1.2.18 on 2026-10-26), about 44k stars.[1]
Cons
- User reportsUsers describe a steep learning curve for people not used to graphs, and say tracing a mid-graph production failure means building your own observability.[1][2]
- User reportsDurability edge cases: long tool calls (about 180s or more) silently re-execute from a checkpoint on LangGraph Cloud, and sync durability doesn't enforce write ordering.[1][2]
- User reportsCancelling a run loses streamed state that hasn't been checkpointed yet, and the Postgres checkpointer has a long-open SSL error.[1][2]
- User reportsDocs lag the fast-moving API, and breaking changes are a recurring complaint.[1][2]
- User reportsSeveral teams outgrow it and write their own loop, keeping only the graph idea.[1]
- Best for
- Long-running, stateful, resumable workflows with human approval steps where you want explicit control flow.
- Watch out for
- Make tool calls idempotent: a checkpoint replay can run them again.
Assessed Oct 29, 2026 · adversarially reviewedSuggest a correction
People
LangChain
Co-founder & CEO
Evals & benchmarks · Claude Code
Engineer, LangGraph
Multi-agent orchestration
Engineer, LangChain
Multi-agent orchestration · Context engineering & memory
Activity
- Last push
- · yesterday
- Latest release
- 1.2.24
· yesterday - Last 90 days
- 20+ releases
1.2.24langgraph==1.2.24
· yesterdayChanges since 1.2.23 release(langgraph): 1.2.24 (#9752) fix(langgraph): handle REMOVE_ALL_MESSAGES in _messages_delta_reducer and None checkpoints in DeltaChannel (#9735)
1.2.23langgraph==1.2.23
· 2d agoChanges since 1.2.22 release(langgraph): 1.2.23 (#9742) release(prebuilt): 1.1.3 (#9741) fix(langgraph,prebuilt): give Send-dispatched tool calls the same state as an edge (#9728) chore(deps): bump opentelemetry-api from 1.45.1 to 1.45.2 in /libs/langgraph (#9696)…
prebuilt==1.1.3langgraph-prebuilt==1.1.3
· 2d agoChanges since prebuilt==1.1.2 release(prebuilt): 1.1.3 (#9741) fix(langgraph,prebuilt): give Send-dispatched tool calls the same state as an edge (#9728) fix(prebuilt): make wrap_tool_call state overrides reach the tool (#9730)
1.2.22langgraph==1.2.22
· 7d agoChanges since 1.2.21 release(langgraph): 1.2.22 (#9702) release(checkpoint-postgres): 3.2.1 (#9700) release(prebuilt): 1.1.2 (#9701) release(checkpoint): 4.3.2 (#9699) fix(langgraph): stream a child graph on its own thread to a caller streaming subgraphs (#9698)…
checkpointpostgres==3.2.1langgraph-checkpoint-postgres==3.2.1
· 7d agoChanges since checkpointpostgres==3.2.0 release(checkpoint-postgres): 3.2.1 (#9700) release(checkpoint): 4.3.2 (#9699) perf(checkpoint-postgres): run a lone store read outside a pipeline (#9635) perf(checkpoint-postgres): skip the async saver's lock for connection pools (#9638)…
Checked against GitHub on Nov 21, 2026. Release notes are shortened to plain text.
The team behind it
San Francisco, CA · Est. 2022
The de-facto standard toolkit for building LLM applications and agents.
LangChain makes the LangChain framework and LangGraph, the low-level orchestration standard for stateful, controllable agents, plus the LangSmith platform for tracing, evaluating and monitoring them in production.
~$75M raised
In the news
Community · Hacker News · Sep 25, 2026
Hello; I was working on optimizing some CUDA kernels and I thought may be it is a good oppurtunity learn langgraph as well. I created a simple C++ CUDA Test Harness and handed that to AI agents. They can run kernels, get benchmarks, and even can profile via nsight
In research (7)
- Agent-Native by Design: A Programming Language for Agentic ApplicationsVerdict: Incremental
Nov 16, 2026 · Baseline framework and the paper's main motivating example of duplicated routing declarations.
- Evaluating Authorization and Trust Boundaries in Agentic Orchestration FrameworksVerdict: Incremental
Nov 16, 2026 · Evaluated framework (LangGraph 1.1.10); graph state and routing hold the authorization checks.
- EREBOTS 4.0: A framework for conversational assistants with A2A agents and MCP toolsVerdict: Incremental
Oct 26, 2026 · Cited as an alternative orchestration toolkit that leaves identity and user binding to the application.
- I-HITL: Inalienable Human-in-the-LoopVerdict: Important
Oct 26, 2026 · LangGraph interrupts are cited as the kind of in-harness HITL approval gate that an agent controlling the platform can route around.
- OpenDraco: A Platform for Engineering and Experimenting with Configurable Multi-Agent Architectures for Software EngineeringVerdict: Incremental
Oct 23, 2026 · OpenDraco compiles its JSON MAS specs into executable LangGraph workflows.
- Tool-Calling LLM Agents Under Realistic MCP Faults: Scale Does Not Purchase Fault ToleranceVerdict: Useful
Oct 23, 2026 · The LangChain stack runs on LangGraph, and its 25-superstep default sets that stack's loop bound.
Showing the 6 newest of 7 papers.
Neighbours on the workbench
TypeScript · ≈395k stars · MIT
openclaw/openclaw
The viral open-source personal AI assistant that really does things.
Maintained by OpenClaw Foundation
v2026.11.6releasedTypeScript · ≈285k stars · MIT
deepseek-ai/deepseek-harness
Everything is a Plugin — DeepSeek's open agent harness.
Maintained by DeepSeek
dsh-v0.2.2-rc.5 prereleasedTypeScript · ≈212k stars · Sustainable Use License (fair-code; GitHub reports NOASSERTION)
n8n-io/n8n
Fair-code workflow automation platform with native AI capabilities.
Maintained by n8n-io
n8n@1.123.100releasedConnections
Relations from the Harness Hub knowledge graph, as charted in the State of Harness reports.
Maintained by · 1
- LangChainStartup
Built on it · 3
- deepagentsjsProject
- LangChain SkillsProject
- VoltAgentStartup
Mentioned in · 4
- Build AI Research Agents with LangGraph: React, Planning, and MemoryEvent · 2026-12-19
- Evaluating Authorization and Trust Boundaries in Agentic Orchestration FrameworksPaper · 2026-11-16
- OpenDraco: A Platform for Engineering and Experimenting with Configurable Multi-Agent Architectures for Software EngineeringPaper · 2026-10-23
- Interrupt 2026Event · 2026-10-13
