#Bm25
Showing 37 of 37 repositories tagged #bm25, ranked by stars
Simple, Elastic-quality search for Postgres
The AI-native database built for LLM applications, providing incredibly fast hybrid search of dense vector, sparse vector, tensor (multi-vector), and full-text.
AdalFlow: The library to build & auto-optimize LLM applications.
A modular Agentic RAG built with LangGraph โ learn Retrieval-Augmented Generation Agents in minutes.
SeekStorm: vector & lexical search - in-process library & multi-tenancy server, in Rust.
Fast BM25 search in Python, powered by Numpy and Numba
Developer friendly Natural Language Processing โจ
๐ง ็บฏๅ็ Python ๅฎ็ฐ็ RAG ๆกๆถ | FAISS + BM25 ๆททๅๆฃ็ดข | ๆฏๆ Ollama / SiliconFlow | ้ๅๆฐๆๅ ฅ้จๅญฆไน
Similarities: a toolkit for similarity calculation and semantic search. ็ธไผผๅบฆ่ฎก็ฎใๅน้ ๆ็ดขๅทฅๅ ทๅ ๏ผๆฏๆไบฟ็บงๆฐๆฎๆๆๆใๆๆๅพใๅพๆๅพ๏ผpython3ๅผๅ๏ผๅผ็ฎฑๅณ็จใ
Easy-to-Use RAG Framework; CCF AIOps International Challenge 2024 Top3 Solution; CCF AIOps ๅฝ้ ๆๆ่ต 2024 ๅญฃๅๆนๆก
Rust-powered code intelligence CLI for AI coding agents. Builds call graphs and hybrid semantic search indexes (Dense + Sparse + RRF + Reranker) across 7 languages. Ships as native MCP tools for Claude Code and Codex CLI.
Local-first persistent memory for AI coding agents (Claude Code, Cursor, Codex) over MCP. Decisions, lessons and facts live in one SQLite file on your disk. Offline, multilingual.
Supercharge AI Agents, Safely
Local-first, zero-key semantic code search for large and custom codebases โ hybrid vector + keyword retrieval with symbol-aware chunking. Usable as a CLI, Python library, REST API, or web UI.
VectorRAG.Net is a .NET-native high-performance vector database library for semantic search and RAG (Retrieval-Augmented Generation). Core search is based on Random Hyperplane LSH candidate generation with exact rerank by dot/cosine.
Claude Code harness for AI coding agents. Hybrid-RAG librarian picks relevant rules in 0.59 ms p95 (BM25 + vector + graph traversal); process keeper blocks writes until plan + tests approved. 276 rules / 30 mandatory across 12 domains, 6 cross-language analyzers. Neo4j + Tantivy + hnswlib + ONNX.
Fast, AI-agent-native code search in Rust โ hybrid BM25 + semantic, Tree-sitter AST chunking, dependency & impact analysis. Drop-in replacement for grep/cat/read/ls in Claude Code, Codex, Cursor, Aider, OpenHands.
Self-hosted RAG search engine โ 34 formats, BM25+hybrid search, multi-LLM (Gemini/OpenAI/Claude/Ollama), FastAPI + Docker, production-ready in 3 min
Emacs package that helps org-mode users (re)discover similar documents
Local code search combining BM25, vector similarity, and cross-encoder reranking. Parses 60+ languages with tree-sitter, runs entirely offline, and returns structured results with file paths, line ranges, and symbol metadata. Built in Rust.
Fast transcript search for humans & agents. Supports Claude Code, Codex CLI & OpenCode
Django BM25 full-text search using PostgreSQL - a lightweight Elasticsearch alternative
State aware knowledge compression, ingestion, and hybrid retrieval engine. Zero dependencies. Sub-100ms queries.
Multi-repo semantic code search MCP server in Rust โ hybrid vector + BM25 retrieval, tree-sitter AST chunking, fully offline. For OpenCode, Claude Code, Cursor, and any MCP client.
MCP server with local vector search for your codebase. Smart indexing, semantic search, Git history โ all offline.
Fast search engine on object storage, with full text search, vectors, and SQL, natively on Parquet.
DocChat is an AI-powered Multi-Agent RAG system using Docling for structured document parsing and BM25 + vector search retrievers to retrieve fact-checked answers from PDFs, DOCX, and text files, preventing hallucinations. ๐
Semantic code search plugin for Claude Code using hybrid vector search + BM25. 98% accuracy, 5x faster than grep.
MCP-native code retrieval for AI agents โ 84-88% fewer read tokens, BM25F + semantic search, AST chunks, session dedup
โก Cut Claude token usage by 90%+ โ free, open-source, local-first context compression for Claude Code. Hybrid RAG (BM25 + ONNX vectors), AST chunking, reranking. No API needed.
The ultimate memory backend for OpenClaw and Claude Code. Persistent conversation memory with LLM fact extraction, foundation-model reranking, and 77.7% LoCoMo accuracy. MCP native, REST API, self-hosted.
Sentence-Transformers Information Retrieval example on Chinese
Agentic memory built on Postgres
Hybrid RAG Worker on Cloudflare Edge โ Vector + BM25 search, metadata filtering, multimodal vision, MCP server, and intelligent query routing. ~400ms p99.
Find traders actively seeking custom bots, AI agents, or developers. Aggregates Reddit, Twitter, GitHub & Telegramโscores for buying intent (BM25), classifies HOT/WARM/COLD. Free-first: no API keys required.
Unsupervised tableQA and databaseQA on chinese finance question and tabular data
Hybrid RAG (DuckDB vector + BM25 + RRF + recency/keyword priors + optional cross-encoder rerank) as an installable library + CLI.