DeerFlow
DeerFlow is an open-source super agent harness that orchestrates sub-agents, memory, and sandboxes to perform complex tasks, powered by extensible skills. It is designed for developers who want to build and deploy autonomous AI agents that can research, code, and execute tasks in a sandboxed environment. It solves the problem of coordinating multiple AI components into a reliable, production-ready agent system.
🎬 DeerFlow: FULLY FREE Local DEEP Research Agent - Powerful AI Agent Can Do Anything! · WorldofAI
✨ Key features
- Orchestrates sub-agents, memory, and sandboxes for complex tasks.
- Extensible skills and tools, including Claude Code integration.
- Session goals and manual context compaction for efficient workflows.
- Long-term memory and context engineering for improved performance.
- Supports multiple LLM providers and models, including local vLLM.
- Docker and local development options with hot-reload.
🎯 Use cases
- Deep research and report generation.
- Automated coding and repository tasks via sub-agents.
- Building autonomous agents that use web search and file operations.
- Prototyping and benchmarking agent ideas with LLM Space.
- Scheduled tasks and long-running server deployments.
📦 Installation
🧰 Requirements: Requires Python, Docker (optional but recommended), and API keys for LLM providers (e.g., OpenAI, Anthropic) and optional web search (e.g., Tavily).
git clone https://github.com/bytedance/deer-flow.git
cd deer-flow
make setup
For Docker:
make docker-init
make docker-start
See README for full details.
🚀 Usage
make setup
make dev
Or with Docker:
make docker-start
Then interact via the TUI or API.
⚠️ Good to know
DeerFlow 2.0 is a ground-up rewrite and shares no code with v1; active development is on 2.0, and v1 is maintained on a separate branch.
❓ FAQ
What is DeerFlow?
DeerFlow is an open-source super agent harness that orchestrates sub-agents, memory, and sandboxes to do almost anything, powered by extensible skills.
How do I install DeerFlow?
Clone the repository and run make setup to launch an interactive wizard that configures your LLM provider and preferences. For Docker, use make docker-init and make docker-start.
What are the recommended models?
The README recommends using Doubao-Seed-2.0-Code, DeepSeek v3.2, and Kimi 2.5 to run DeerFlow.
Is DeerFlow 2.0 compatible with v1?
No, DeerFlow 2.0 is a ground-up rewrite and shares no code with v1. The original framework is maintained on the 1.x branch.
📊 Repository
🤖 Overview, features, install steps and FAQ were generated from the project's README on Sep 4, 2026. Always check the original source before running commands.