DeepAgents

DeepAgents is an open-source Python library from LangChain AI that enables developers to build deep research-style agentic workflows using large language models. It provides a structured framework for orchestrating multi-agent pipelines that can plan, search, reason, and synthesize information across multiple steps. Built on top of LangChain's ecosystem, it is designed for developers who want to create autonomous research agents with minimal boilerplate.

• Multi-agent orchestration — Build and coordinate multiple specialized agents that collaborate to complete complex research tasks.
• Deep research workflows — Implement iterative search-and-synthesize loops inspired by deep research paradigms for thorough information gathering.
• LangChain integration — Leverages the full LangChain ecosystem including LLM providers, tools, and memory modules out of the box.
• Open-source and extensible — Fully open-source under a permissive license; extend or customize agent behavior by modifying the source code directly.
• LLM-agnostic design — Compatible with a wide range of LLM backends supported by LangChain, including OpenAI, Anthropic, and others.
• Tool use support — Agents can be equipped with web search, code execution, and other tools to gather and process information autonomously.
• Python-native — Install via pip and integrate directly into existing Python projects or notebooks with minimal setup.