DeerFlow 2.0 - ByteDance's Open-Source Super Agent Framework

What Is DeerFlow 2.0

DeerFlow 2.0 is an open-source super agent framework from ByteDance, built on a "main agent + 11-layer middleware chain + dynamic sub-agents" architecture, with multi-agent collaboration implemented through LangGraph. The framework comes with 10+ pluggable skills built in, including deep research, data analysis, and audio/video creation, and supports isolated sandbox execution via Docker/K8s. It natively integrates with Feishu, Telegram, and Slack, and can run without a public IP. Version 2.0 upgrades it from a single research tool to a general-purpose agent runtime, supporting complex tasks ranging from minutes to hours and enabling one-click generation of websites, PPTs, comics, and more.

Main Features of DeerFlow 2.0

• Agent Architecture: Adopts a collaboration model in which a main agent handles overall planning while dynamic sub-agents execute in parallel, enabling efficient decomposition and collaborative handling of complex tasks.
• Pluggable Skill System: Comes with more than 10 skill modules built in, including deep research, data analysis, PPT generation, and web design, and supports users in customizing and extending new capabilities through the MCP interface and Python functions.
• Isolated Sandbox Execution: Provides three operating modes: local, Docker, and Kubernetes. Based on ByteDance's AIO Sandbox, it creates an independent container environment for each task to ensure execution security and resource isolation.
• Context Engineering: Through mechanisms such as automatic summary compression, external storage of intermediate results, and subtask throttling, it systematically solves the problem of insufficient context windows in long-running tasks.
• Long-Term Memory: Supports persistent storage of user profiles, preferences, and accumulated knowledge across sessions, with all data stored locally and fully under the user's control.
• IM Channel Integration: Natively integrates with the three major instant messaging platforms Feishu, Telegram, and Slack, allowing interaction with the agent through commands without the need for a public IP.
• Multi-Model Compatibility: Supports mainstream large models such as OpenAI, Gemini, DeepSeek, and Doubao Seed, enabling flexible switching between different underlying capabilities through a unified configuration interface.

Key Information and Usage Requirements for DeerFlow 2.0

• Project Background: An open-source super agent framework from ByteDance, released under the MIT License.
• Release Date: Version 2.0 was officially released on February 28, 2026, and topped the GitHub Trending list on the day of release.
• Core Positioning: Fully upgraded from the deep research tool of version 1.0 to a general-purpose agent runtime (Super Agent Harness).
• Task Capabilities: Supports handling complex multi-step tasks ranging from minutes to hours, covering scenarios such as research, coding, and creation.
• Python Version: Requires Python 3.12 or higher to be installed on the system.
• Node.js Version: Requires Node.js 22 or higher to be installed on the system.
• Package Manager: Requires the installation of pnpm and uv as frontend and backend package management tools.

How to Use DeerFlow 2.0

• Clone the Repository: Run git clone https://github.com/bytedance/deer-flow.git to clone the project locally, then enter the project directory cd deer-flow。
• Generate Configuration: Run make config to automatically generate the config.yaml and .env configuration files.
• Configure the Model: Edit config.yaml file, add the required models (such as GPT-4, Gemini, DeepSeek, Doubao, etc.), and set parameters such as the model name, API type, and key variables.
• Set the key: In the .env file, enter the API key for each service provider, such as OPENAI_API_KEY=your-key。
• Choose a deployment method
  • Docker method: Run make docker-init to pull the sandbox image, then run make docker-start to start the service.
  • Local method: Run make check to check dependencies, then run make install to install dependencies, and finally run make dev to start the service.
• Access and use: Open a browser and visit http://localhost:2026 to enter the Web interface, then enter a task instruction to start using it.
• IM channel integration: In config.yaml , enable the Feishu, Telegram, or Slack configuration and set parameters such as the corresponding platform's App ID and key, so you can interact with DeerFlow through commands in the chat software.

DeerFlow 2.0 project address

• Project official website：https://deerflow.tech
• GitHub repository：https://github.com/bytedance/deer-flow

Comparison of DeerFlow 2.0 with similar competing products

Application scenarios of DeerFlow 2.0

• Deep research: Automatically collects multi-source information and generates research reports, suitable for complex information processing tasks such as academic research, competitive analysis, and industry research.
• Web development: Generates complete deliverable website pages in one click, automating the entire process from UI design to front-end code, such as football league official websites, 3D interactive weather interfaces, etc.
• Content creation: Transforms complex concepts into child-friendly educational comics, or automatically generates multimedia materials such as PPTs, podcast scripts, and video content.
• Data analysis: Automatically execute data-driven tasks such as data cleaning, visualization chart generation, and business intelligence report writing.