LangGraph
Open-source framework for building stateful, graph-based LLM agent workflows, from the LangChain team.
HN commenters use it to orchestrate agent systems as state machines, including in production,1234 valuing its modularity, explicit control over logic, and features like time travel and human-in-the-loop interruptions.5678 The most common caveat is that agents built this way can be annoying to make reliable,9 and some now prefer Claude Code or plain generated scripts over a dedicated agent framework.1011
Why people recommend it
Agent orchestration · 3
Commenters build multi-agent and multi-stage systems on it as the orchestration layer.
“CoreSmith is built on LangGraph.”
“LangGraph for the agent orchestration, async pipelines throughout.”
Show 1 more
“We use LangGraph to implement the State Machine pattern rather than a linear Chain.”
Control and reliability · 3
Passing explicit state through the graph is seen as a reliable way to control agent logic.
“having that text state to pass into something like LangGraph is the only way to reliably control the logic”
“it seems to provide a lot of infrastructure for highly reliable and interactive agents”
Show 1 more
“I switched from a similar parallelized setup to LangGraph precisely because the merge conflicts and redundant reasoning steps were killing my margins.”
Low-level features · 1
Offers time travel, human-in-the-loop interruptions, and flexibility on the paradigm.
“We are using LangGraph, which is quite low level, but has useful features like time travel, human in the loop capabilities (interruptions)”
Modularity · 1
Its modular design makes individual components easier to test.
“For the agentic architecture we picked LangGraph — the modularity makes components easier to test.”
Model agnostic · 1
Described as model agnostic.
“I think the OpenAI Agents SDK is more model agnostic, LangGraph too.”
Caveats and dissent
Reliability effort · 1
One commenter found getting LangGraph agents to behave reliably annoying and tricky.
“Getting those agent systems to behave reliably was often annoying and tricky.”
Coding agents alternative · 2
Some commenters trust Claude Code more or prefer generating plain scripts instead of using an agent framework.
“we use langgraph in production”
“I usually stay away from systems which already make it seem as if building agents is a ritual.”
Low level · 1
Described as quite low level.
“We are using LangGraph, which is quite low level, but has useful features like time travel, human in the loop capabilities (interruptions)”
Often compared with
Who recommends it in their own words
Threads where it came up
- Ask HN: Freelancer? Seeking freelancer? (September 2026) 1 mention
- Elon House 1 mention
- Ask HN: Who wants to be hired? (August 2026) 1 mention
- Qwen3.8-Max: A New Bar for Coding and Cowork 1 mention
- Towards a Harness That Can Do Anything 1 mention
- item 48873362 1 mention
Also written as langgraph. Data: langgraph.json. Quotes are verbatim; the verdict and theme labels are written by a model from these comments, and every clause links to its sources.
All comments 18
“Agents: LangGraph, LangChain”
“I used langgraph + ws + all in cloudflare + 2 llm providers”
“Python, FastAPI, LangGraph, MCP, RAG, Postgres/pgvector, Redis, Celery, Docker”
“someone who wanted me to start over in LangGraph”
“I usually stay away from systems which already make it seem as if building agents is a ritual.”
“Getting those agent systems to behave reliably was often annoying and tricky.”
“For the agentic architecture we picked LangGraph — the modularity makes components easier to test.”
“CoreSmith is built on LangGraph.”
“We are using LangGraph, which is quite low level, but has useful features like time travel, human in the loop capabilities (interruptions)”
“as an alternative to LangGraph”
“LangGraph for the agent orchestration, async pipelines throughout.”
“I used AI agent for my last langgraph solo project”
“I think the OpenAI Agents SDK is more model agnostic, LangGraph too.”
“having that text state to pass into something like LangGraph is the only way to reliably control the logic”
“we use langgraph in production”
“I switched from a similar parallelized setup to LangGraph precisely because the merge conflicts and redundant reasoning steps were killing my margins.”
“it seems to provide a lot of infrastructure for highly reliable and interactive agents”
“We use LangGraph to implement the State Machine pattern rather than a linear Chain.”