bytedance/deer-flow

What is it?

Key features

• 01Sub-agents that split and run parts of a long task in parallel.
• 02Isolated sandboxes for file operations and code execution.
• 03Long-term memory that persists across sessions.
• 04Extensible skills and tools, plus an MCP server.
• 05Configurable LLM providers via config.yaml, including OpenAI-compatible endpoints and Claude.
• 06Web search and crawling integrations with multiple providers.

Use cases

• Running multi-step web research that produces a written report.
• Delegating coding or refactoring tasks to an agent working in a sandbox.
• Building a self-hosted agent workspace that connects to chat channels.
• Experimenting with multi-agent workflows on top of LangGraph.

Quick start

1. 1

   Clone the DeerFlow repository

   git clone https://github.com/bytedance/deer-flow.git
2. 2

   Enter the project directory

   cd deer-flow
3. 3

   Run the setup wizard to choose an LLM provider and generate config

   make setup

Things to know

• !Requires an LLM provider and API key, which the setup wizard asks you to configure.
• !Needs Python 3.12+ and Node.js 22+ for the local development setup.
• !The README includes a security notice: improper deployment can introduce risks, and gateway admin access is equivalent to code execution.

Topics

FAQ

What is deer-flow?

Open-source long-horizon super agent harness that orchestrates sub-agents, memory, and sandboxes to research, code, and create across tasks lasting minutes to hours. DeerFlow (Deep Exploration and Efficient Research Flow) is a super agent harness written in Python with a TypeScript frontend. It coordinates sub-agents, long-term memory, sandboxed execution environments, and extensible skills so one agent session can handle tasks that run from several minutes to several hours.

Is deer-flow free to use? What license does it have?

deer-flow is open source under the MIT license. Check the license terms before using it in commercial products.

Is deer-flow still maintained?

Actively maintained. The last commit to deer-flow was on Oct 20, 2026.

What can I use deer-flow for?

Running multi-step web research that produces a written report. Delegating coding or refactoring tasks to an agent working in a sandbox. Building a self-hosted agent workspace that connects to chat channels. Experimenting with multi-agent workflows on top of LangGraph.

How do I install deer-flow?

Clone the DeerFlow repository: git clone https://github.com/bytedance/deer-flow.git Enter the project directory: cd deer-flow Run the setup wizard to choose an LLM provider and generate config: make setup

Where can I get deer-flow?

The source code and installation instructions are on GitHub: https://github.com/bytedance/deer-flow