Chapter 6: Agents

The reason why LangChain can be used to build any sort of complex app around LLMs is because of agents. Agents enable the development of a diverse range of customized apps, from being as simple as a Q&A bot to as complex as AutoGPT AI agent. We will first start by explaining what an agent is.


An agent can be taken as an extension of chains where we can equip the LLM with a variety of 3rd party tools or even APIs like Google Search and get our task done by following multiple actions. If you remember, even chains can handle 3rd party tools and follow a sequence of actions. Then what’s the difference?

6.1 How are Agents different from Chains?

The actions in a chain are hardcoded and need to follow a predefined sequence but in ...

Get LangChain in your Pocket now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.