DeerFlow
ActiveDescription
An open-source long-horizon SuperAgent harness by ByteDance that researches, codes, and creates with sandboxes, memories, tools, skills, subagents and message gateway for complex tasks.
Key Features
- Super Agent harness - orchestrates sub-agents, memory and sandboxes for complex long-horizon tasks from minutes to hours
- Extensible skills system - modular skill-based architecture for extending agent capabilities with custom skills
- Sandboxed code execution - built-in secure sandbox for agents to safely execute code and file operations in isolation
- Long-term memory - cross-session memory persistence for maintaining context coherence across interactions
- Context engineering - sophisticated context management and orchestration for information flow in long tasks
- Multi-model support - supports OpenAI, Anthropic, Gemini, OpenRouter, vLLM and multiple LLM providers
Use Cases
💡 Deep research automation where agents autonomously perform multi-round search, reading, analysis and report writing
💡 Complex coding projects with agents writing, testing and iterating code in sandboxes until features are complete
💡 Content creation workflows orchestrating multiple sub-agents for end-to-end research-to-draft production
💡 Enterprise knowledge management building organizational knowledge agents with continuous learning via long-term memory
💡 Data analysis report generation with agents autonomously performing data processing, statistical analysis and visualization
Strengths & Limitations
✅ Strengths
- • Actively maintained, recent updates
- • High community interest (81.3k stars)
- • Permissive open-source license (MIT)
Categories
Quick Start
# Clone the project
git clone https://github.com/bytedance/deer-flow.git
cd deer-flow
# Run the interactive setup wizard
make setup
# Start with Docker
make docker-up
# Or local development mode
make dev
# Verify setup
make doctor