# infiniflow/infinity

The AI-native database built for LLM applications, providing incredibly fast hybrid search of dense vector, sparse vector, tensor (multi-vector), and full-text.

Repository: https://github.com/infiniflow/infinity
Canonical: https://ross.abutalabs.com/products/infinity
Homepage: https://infiniflow.org
Language: C++
License: Apache-2.0
License Family: permissive
Topics: ai-native, nearest-neighbor-search, rag, vector-search, information-retrival, bm25, embedding, full-text-search, search-engine, cpp20-modules, approximate-nearest-neighbor-search, vector-database, hybrid-search, hnsw, cpp20, tensor-database, vector, vectordatabase, multi-vector
Last push: 2026-08-17T13:43:09+00:00

## Health v2 (maintenance only)
Score: 98/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 98, release rhythm 96, longevity 100
- inputs: {"age_days": 1507, "days_push": 16, "days_rel": 27, "gap_med": 6, "n_releases_24m": 50}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4683, forks 438 (observed 2026-08-28T04:08:56.943056+00:00)

## What it is
Infinity is an AI-native database built for LLM applications, offering high-performance hybrid search across dense vectors, sparse vectors, tensors, and full text with filtering and reranking (RRF, weighted sum, ColBERT). It is a single-binary server with a Python SDK, deployable via Docker or embedded in Python.

## Use cases
- build a RAG pipeline with hybrid vector and full-text search
- store and query embeddings for semantic search
- run BM25 full-text search over millions of documents
- combine dense, sparse, and tensor search with rerankers like ColBERT
- self-host a vector database for a chatbot or QA app
- filter vector search results by structured data

## When to choose
- you need fast hybrid search combining vectors and full text in one database
- you want a dependency-free single-binary vector database with a Python API
- you need reranking (RRF, weighted sum, ColBERT) built into the query engine
- you are building LLM/RAG applications needing sub-millisecond retrieval

## When to avoid
- you only need a general-purpose relational database without vector search
- your hardware lacks AVX2 support
- you need a fully managed cloud database service
- you require mature multi-node distributed clustering

## Facets
- artifact type: service
- maturity: active
- function: vector-database, search-engine, database, rag
- domain: databases, large-language-models, artificial-intelligence
- platform: windows, self-hosted, python, cpp
- tags: hybrid-search, full-text-search, bm25, hnsw, sparse-vectors, tensor-search, reranking, colbert, embeddings, ai-native-database, search, retrieval-augmented-generation, linux, macos, docker

## Member repositories
- infiniflow/infinity (main) score 98

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:56.943056+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 2026-08-29T18:19:18.618902+00:00, confidence not recorded.
  - readme: https://github.com/infiniflow/infinity (fetched 2026-08-28T04:08:56.943056+00:00, sha 5cf8264d813e)
  - homepage: https://infiniflow.org (fetched 2026-08-29T09:03:28.703366+00:00, sha cba16db49a4c)
  - site_page: https://infiniflow.org/docs/category/get-started (fetched 2026-08-29T09:03:28.705807+00:00, sha a06a818e548e)
  - site_page: https://infiniflow.org/docs (fetched 2026-08-29T09:03:28.707421+00:00, sha ed36803599e6)
  - site_page: https://infiniflow.org/docs/dev (fetched 2026-08-29T09:03:28.709077+00:00, sha 8dd003516fa7)
- Data as of 2026-08-30T08:39:29.467469+00:00.