# OSU-NLP-Group/HippoRAG

[NeurIPS'24] HippoRAG is a novel RAG framework inspired by human long-term memory that enables LLMs to continuously integrate knowledge across external documents. RAG + Knowledge Graphs + Personalized PageRank.

Repository: https://github.com/OSU-NLP-Group/HippoRAG
Canonical: https://ross.abutalabs.com/products/hipporag
Homepage: https://arxiv.org/abs/2405.14831
Language: Python
License: MIT
License Family: permissive
Last push: 2026-08-23T14:35:10+00:00

## Health v2 (maintenance only)
Score: 59/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 8, longevity 59
- inputs: {"age_days": 832, "days_push": 10, "days_rel": 552, "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 3967, forks 419 (observed 2026-08-28T04:08:30.717441+00:00)

## What it is
HippoRAG is a Python RAG framework inspired by human long-term memory that combines LLMs, knowledge graphs, and Personalized PageRank to integrate knowledge across documents. It improves multi-hop retrieval and sense-making while remaining cheaper and faster than iterative or other graph-based RAG approaches.

## Use cases
- answer multi-hop questions across documents
- build a knowledge-graph-based RAG pipeline
- give LLMs long-term memory over a document corpus
- cheaper alternative to GraphRAG for offline indexing
- improve retrieval accuracy for complex question answering
- continually integrate new knowledge without retraining

## When to choose
- you need multi-hop or associative retrieval over a large corpus
- you want graph-based RAG with lower indexing cost than GraphRAG or LightRAG
- you're doing research on memory-augmented LLMs and want a published, benchmarked method

## When to avoid
- you need a simple plug-and-play vector RAG with minimal setup
- your queries are single-hop factual lookups where standard RAG suffices
- you can't run LLM calls or GPU inference for offline indexing

## Facets
- artifact type: library
- maturity: active
- function: rag, search-engine, machine-learning, llm-inference, nlp
- domain: large-language-models, artificial-intelligence, machine-learning
- platform: python
- tags: knowledge-graph, personalized-pagerank, multi-hop-qa, long-term-memory, graphrag, neurips, retrieval-augmented-generation, natural-language-processing, linux, macos, gpu

## Member repositories
- OSU-NLP-Group/HippoRAG (main) score 59

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:30.717441+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 2026-08-29T18:24:23.840867+00:00, confidence not recorded.
  - readme: https://github.com/OSU-NLP-Group/HippoRAG (fetched 2026-08-28T04:08:30.717441+00:00, sha 9ac2bc70038f)
  - homepage: https://arxiv.org/abs/2405.14831 (fetched 2026-08-29T09:17:54.904852+00:00, sha 30018fb7e886)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T09:17:54.913915+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T09:17:54.917499+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T09:17:54.919418+00:00, sha a1f16f915a9a)
  - registry_pypi: https://pypi.org/pypi/hipporag/json (fetched 2026-08-29T09:17:54.921256+00:00, sha e3e0c2c68b14)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T09:17:54.915723+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.