# bytedance/deer-flow

An open-source long-horizon SuperAgent harness that researches, codes, and creates. With the help of sandboxes, memories, tools, skill, subagents and message gateway, it handles different levels of tasks that could take minutes to hours.

Repository: https://github.com/bytedance/deer-flow
Canonical: https://ross.abutalabs.com/products/deer-flow
Homepage: https://deerflow.tech
Language: Python
License: MIT
License Family: permissive
Topics: agent, agentic, agentic-framework, agentic-workflow, ai, ai-agents, deep-research, langchain, langgraph, llm, multi-agent, nodejs, podcast, python, langmanus, typescript, harness, superagent
Last push: 2026-08-26T07:04:01+00:00

## Health v2 (maintenance only)
Score: 72/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 58, longevity 34
- inputs: {"age_days": 483, "days_push": 7, "days_rel": 69, "gap_med": null, "n_releases_24m": 1}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 80958, forks 11142 (observed 2026-08-28T04:12:21.885259+00:00)

## What it is
DeerFlow is an open-source SuperAgent harness from ByteDance that orchestrates a lead agent with sub-agents, memory, tools, skills, and Docker-based sandboxes to handle long-horizon research, coding, and content-creation tasks. It consists of a core runtime SDK (Harness) plus a reference application layer for deployment and end-user workflows.

## Use cases
- run deep research reports on any topic
- build a multi-agent system with memory and subagents
- execute long-running coding tasks in a sandboxed environment
- generate podcasts or videos from research
- automate multi-step agentic workflows that take minutes to hours
- extend an agent with custom skills and MCP tools
- self-host an AI agent workspace with a web UI

## When to choose
- you need a full agent harness with sandboxes, memory, and subagent orchestration out of the box
- you want deep research plus code execution and content generation in one system
- you want an MIT-licensed, self-hostable agent framework with an optional reference app
- you want to extend agent capabilities via skills and MCP integrations

## When to avoid
- you need a lightweight single-agent chatbot without orchestration overhead
- you want a fully managed cloud agent service rather than self-hosting
- your stack cannot run Python 3.12+ and Node.js 22+ together
- you need a simple deep-research-only tool and prefer the lighter 1.x branch

## Facets
- artifact type: framework
- maturity: active
- function: agent-framework, rag, llm-inference, web-scraping, chatbot, mcp, sdk
- domain: artificial-intelligence, large-language-models, developer-tools
- platform: python, self-hosted, cross-platform
- tags: superagent, deep-research, multi-agent, langgraph, langchain, agent-harness, sandboxed-code-execution, subagents, agent-memory, skills, podcast-generation, ai-agents, retrieval-augmented-generation, natural-language-processing, nodejs, docker

## Member repositories
- bytedance/deer-flow (main) score 72

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:12:21.885259+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:13:56.385896+00:00, confidence not recorded.
  - readme: https://github.com/bytedance/deer-flow (fetched 2026-08-28T04:12:21.885259+00:00, sha ad18ec175b58)
  - homepage: https://deerflow.tech (fetched 2026-08-28T17:41:07.610528+00:00, sha f864c4f8003b)
  - site_page: https://deerflow.tech/en/docs (fetched 2026-08-28T17:41:07.619099+00:00, sha ade84af2d700)
- Data as of 2026-08-30T08:39:29.467469+00:00.