# TencentCloudADP/youtu-graphrag

[ICLR 2026] Youtu-GraphRAG: Vertically Unified Agents for Graph Retrieval-Augmented Complex Reasoning

Repository: https://github.com/TencentCloudADP/youtu-graphrag
Canonical: https://ross.abutalabs.com/products/youtu-graphrag
Homepage: https://arxiv.org/abs/2508.19855
Language: Python
License: NOASSERTION
License Family: other
Topics: agent, graph, graphrag, llm, rag
Last push: 2026-02-26T12:57:55+00:00

## Health v2 (maintenance only)
Score: 57/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 69, release rhythm 60, longevity 26
- inputs: {"age_days": 366, "days_push": 188, "days_rel": 188, "gap_med": 80.5, "n_releases_24m": 3}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1253, forks 183 (observed 2026-08-28T04:04:08.554868+00:00)

## What it is
Youtu-GraphRAG is a Python framework for graph-based retrieval-augmented generation that unifies graph schema construction, community detection, and agentic retrieval for complex multi-hop reasoning over knowledge graphs. It was accepted at ICLR 2026 and offers an interactive interface plus an enterprise edition on Tencent Cloud's ADP platform.

## Use cases
- build a graphrag pipeline over private domain knowledge
- answer multi-hop complex questions with llm reasoning
- summarize and query a knowledge graph built from documents
- transfer graphrag to new domains with minimal schema changes
- reduce token cost in retrieval-augmented generation
- hierarchical knowledge retrieval with community summaries

## When to choose
- you need structured, graph-based RAG with multi-hop reasoning over domain knowledge
- you want domain transfer with minimal schema intervention
- token cost and accuracy over SOTA baselines matter to you

## When to avoid
- you need simple vector-only RAG without graph structure
- you require a permissively licensed production dependency (license is non-standard)
- you want a fully managed service rather than self-hosting

## Facets
- artifact type: framework
- maturity: active
- function: rag, agent-framework, llm-inference, search-engine, nlp
- domain: large-language-models, machine-learning
- platform: python, self-hosted
- tags: graphrag, knowledge-graph, multi-hop-reasoning, community-detection, iclr-2026, tencent, retrieval-augmented-generation, ai-agents, natural-language-processing, docker, web-server

## Member repositories
- TencentCloudADP/youtu-graphrag (main) score 57

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:08.554868+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 2026-08-30T05:07:27.478533+00:00, confidence not recorded.
  - readme: https://github.com/TencentCloudADP/youtu-graphrag (fetched 2026-08-28T04:04:08.554868+00:00, sha 17f8a8775a8e)
  - homepage: https://arxiv.org/abs/2508.19855 (fetched 2026-08-29T12:18:14.030594+00:00, sha 2190981dcc20)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T12:18:14.041008+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T12:18:14.045056+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T12:18:14.047101+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T12:18:14.043265+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.