// rag · infiniflow/infinity

infinity

infinityは、infiniflowが公開しているオープンソースのRAG・知識ベースプロジェクトです（GitHub: infiniflow/infinity）。GitHubスターは4,734、フォークは455で、Apache-2.0ライセンスの下で公開されています。GitHubでの最終更新は2026年9月23日です。

AI-native database for hybrid search across dense, sparse, and tensor vectors plus full text

★ 4.7k スター⑂ 455 フォークApache-2.0更新 2026年9月

GitHubで見る ↗Official site ↗

▦ ビジュアル — このプロジェクトのREADMEには埋め込み可能なスクリーンショットがないため、
ここではプレースホルダーを表示しています。下のガイド自体は完全な内容です。

// プロジェクトの基本情報

プロジェクト基本情報

リポジトリ: infiniflow/infinity
GitHubスター: 4,734
フォーク: 455
ライセンス: Apache-2.0
GitHubでの最終更新: 2026年9月23日
カテゴリ: RAG・知識ベース
掲載タイプ: 完全なインストールガイド
公式サイト: Official site

// 仕組み

仕組み

Plan your deployment
Run the server in Docker (recommended) or from a binary; an x86_64 CPU with AVX2 and Python 3.11+ for the client are expected.
Start the server
Pull and run the Infinity Docker image with a persistent data volume.
Install the client SDK
Install the Python SDK matching the server.
Run your first hybrid search
Create a table with text and vector columns, insert rows, and query with dense matching plus filters.

// 機能

選ばれる理由

Hybrid search in one engine

Combine dense embedding, sparse embedding, tensor, and full-text search with filtering in a single query.

Multiple rerankers

RRF, weighted sum, and ColBERT reranking are supported out of the box.

Rich data types

Strings, numerics, vectors, tensors, and structured data in one database.

Single-binary deployment

No external dependencies; deploy via Docker or binary, or embed it in Python as a module.

Python and HTTP APIs

An intuitive Python API/SDK and an HTTP API for other clients.

Built for RAG workloads

Designed for LLM applications including search, recommenders, Q&A, conversational AI, and copilots.

// 自分で実行

自分で実行してみる

Clone the repository
Clone if you plan to build from source or browse the examples; Docker users can skip ahead.
```
git clone https://github.com/infiniflow/infinity.git
cd infinity
```

Run the server with Docker

Start Infinity with host networking and a persistent data directory, as documented.

sudo mkdir -p /var/infinity && sudo chown -R $USER /var/infinity
docker pull infiniflow/infinity:nightly
docker run -d --name infinity -v /var/infinity/:/var/infinity --ulimit nofile=500000:500000 --network=host infiniflow/infinity:nightly

Install the client and search
Install the Python SDK, then connect on port 23817 and run a dense-vector match.
```
pip install infinity-sdk==0.7.3
```

// 活用例

作れるもの

Power a RAG pipeline that needs vector, keyword, and tensor search in one store
Build semantic product search with full-text fallback and filters
Serve multi-vector (ColBERT-style) retrieval for late-interaction models
Back a copilot or Q&A app that mixes structured filters with embedding search

// よくある質問

よくある質問

Who maintains Infinity?

InfiniFlow, the company also known for RAGFlow; Infinity is their dedicated AI-native database project.

Is Infinity free?

Yes, it is open source under the Apache-2.0 license; you run it on your own infrastructure.

Does it need a paid API key?

No. Infinity is self-hosted software; you generate embeddings with your own models or providers separately.

How does it compare to Milvus or Qdrant?

Those are primarily vector databases; Infinity's pitch is a single engine that natively combines dense, sparse, and tensor search with full-text search and structured data, deployed as one dependency-free binary.

How do I get started?

LibreAIカタログに掲載されたRAG・知識ベースプロジェクトの中で最もスターの多いもの — infinityが合わなかった場合に検討したい、現実的な候補リストです。

MarkItDown★ 193.5k
MarkItDown — Microsoft tool that converts PDFs, Office files, images and more to clean Markdown for LLMs, MIT
Dify★ 157.8k
Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or…
ragflow★ 91.6k
RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG with Agent capabilities to create a superior context…
PaddleOCR★ 90.8k
Lightweight OCR toolkit turning PDFs and images into structured, LLM-ready data in 100+ languages
MinerU★ 82.5k
MinerU — turns PDFs and Office documents into LLM-ready Markdown and JSON for agents and RAG pipelines, Apache-2.0 plus terms

// さらに見る rag

さらに見る： RAG・知識ベース

カタログの中でinfinityとスター数が近いRAG・知識ベースプロジェクト — 前後の棚もそのまま見て回れます。

OpenKB★ 4.8k
OpenKB compiles your documents into a persistent, interlinked wiki knowledge base — no vector DB
ragapp★ 4.4k
RAGApp: deploy a RAG app over your documents with Docker — configure it like a custom GPT
AutoRAG★ 5.1k
An agent that finds anything on your computer — federated search over your files, in place.
ragent★ 4.2k
Ragent (容纳智能体): a production-grade Java Agentic RAG platform with hybrid retrieval and an admin console
code-graph-rag★ 5.2k
RAG for your monorepo: query, understand and edit multi-language codebases with a knowledge graph.
nano-graphrag★ 4.0k
nano-graphrag: a small, hackable GraphRAG implementation in about 1,100 lines — no Microsoft GraphRAG bulk

infinityを入手する

無料でオープン — ソースコードを入手して、リポジトリにスターを付け、作り始めましょう。

GitHubで見る ↗← RAG・知識ベースのツールをさらに見る