Agents
I think of it like this: If you are building a simple agent with custom tools in your app -> langchain create agent If you are simply doing …
I think of it like this: If you are building a simple agent with custom tools in your app -> langchain create agent If you are simply doing some LLM analysis or synthetic data creation -> langchain in
I think of it like this: If you are building a simple agent with custom tools in your app -> langchain create agent If you are simply doing some LLM analysis or synthetic data creation -> langchain invoke If you are building a long horizon agent, something that needs to act autonomously -> deepagents If you have need for something very bespoke and configurable because your software/industry operates at very low levels of scrutiny like biology/finance then you probably need to use langgraph and build your own harness. So in essence if you have a need for your own harness -> langgraph. They all really cover almost every type of agent you would ever want to build at different levels of the stack. langchain, langgraph, and deepagents are our three core open source projects they each occupy a different place in the ecosystem
Related
- Created my own cli agent with LangChain deep agent cli. It's crazy how simple it is to create such agent today.
- deepagents subagents are just tools. when you call a subagent, thats conceptually a function call. this is the simplest mental model for bui…
- That's a quick tour through building and deploying a simple agent with DeepAgents deploy Was just a markdown file, a json file, and toml fil…
Source: Harrison Chase (X) | 2026-08-06