langchain-ai/deepagents
Deep Agents is an agent harness built on langchain and langgraph. Deep Agents are equipped with a planning tool, a filesystem backend, and the ability to spawn subagents - making them well-equipped to handle complex agentic tasks.
Built on LangGraph, it provides automatic context management through conversation summarization and output-to-file routing, plus shell execution with optional sandboxing and MCP (Model Context Protocol) integration for extensible tool ecosystems. The agent returns a compiled LangGraph graph, enabling production features like streaming, persistence, and checkpointing without additional orchestration. Smart defaults include task planning with progress tracking and hierarchical sub-agent delegation with isolated context windows, reducing boilerplate while remaining fully customizable via tools, prompts, and model swapping.
10,665 stars and 2,341,890 monthly downloads. Used by 6 other packages. Actively maintained with 491 commits in the last 30 days. Available on PyPI.
Stars
10,665
Forks
1,705
Language
Python
License
MIT
Category
Last pushed
Mar 13, 2026
Monthly downloads
2,341,890
Commits (30d)
491
Dependencies
5
Reverse dependents
6
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/agents/langchain-ai/deepagents"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Compare
Related agents
kyopark2014/langgraph-agent
It shows case studies of the LangGraph agent.
ArcadeAI/SlackAgent
An assistant for Slack built with Arcade and Langgraph. Interact with Google Calendar, Mail,...
codingforentrepreneurs/django-ai-agent
Learn how to create an AI Agent with Django, LangGraph, and Permit.
Yangyf1125/EasyAgent
A simple agent built on LangGraph
langchain-ai/agents-from-scratch-ts
LangGraph Typescript Agents Notebooks: email, human in the loop, memory