#99th of 15 in Agent frameworks and SDKs
deepagents
Python agent harness with sub-agents, filesystem tools and context management
- Stars
- 30.3k
- License
- MIT
- Last commit
- Oct 2026
- Last release
- Oct 2026
- Language
- Python
17,016 on 24 Mar 202630,255 on 13 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
LangChain is hiring: 100 open roles
Overview
Deep Agents is a Python library whose create_deep_agent() returns a ready-to-run agent built on LangChain's create_agent and LangGraph. It bundles planning, sub-agents with isolated context, a pluggable filesystem and shell backend, context summarization, skills, human-in-the-loop approval and persistent memory. It works with any tool-calling LangChain chat model, including Ollama, vLLM and llama.cpp.
Who it is for: Python developers building long-running, tool-using agents on LangChain
Strengths
- Sub-agents, filesystem, context summarization and skills are bundled and work out of the box
- Any LangChain chat model with tool calling works, including Ollama, vLLM and llama.cpp
- Any LangGraph CompiledStateGraph can be passed in as a sub-agent
- Human-in-the-loop approval of tool calls; tools can be custom functions or MCP servers
Weaknesses
- Opinionated defaults sit on top of LangChain and LangGraph, adding layers to learn
- Trust-the-LLM security model: boundaries must be enforced via tools or sandbox
- Tracing and evaluation are documented around LangSmith, a separate hosted product
- Python only here; the JavaScript version is a separate repository
What it needs
- no GPU
- Needs LangChain, LangGraph, LLM provider or local model server
- Models: OpenAI, Anthropic, Google, Baseten, Fireworks
Also in Agent frameworks and SDKs
See all 15| Rank | Project | Score |
|---|---|---|
| 1 | langchainPython framework for building LLM agents and apps from interoperable components | 88 out of 100 |
| 2 | browser-usePython library that lets an LLM agent control a web browser | 88 out of 100 |
| 3 | crewaiPython framework for multi-agent workflows with role-based Crews and event-driven Flows | 80 out of 100 |
| #4 | openai-agents-pythonPython SDK for multi-agent workflows with handoffs, guardrails and tracing | 77 out of 100 |
| #5 | agnoPython SDK and runtime for building and serving agents behind a REST API | 76 out of 100 |