Hybrid Search Python Packages

flashrank

Lite & Super-fast re-ranking for your search & retrieval pipelines. Supports SoTA Listwise and Pairwise reranking based on LLMs and cross-encoders and more. Created by Prithivi Da, open for PRs & Collaborations.

topk-sdk

High-quality search for AI-native applications.

pylibseekdb

Multi-language client bindings for SeekDB (Python, Node.js), powered by libseekdb

mnemo-mcp

Persistent AI memory with hybrid search and embedded sync. Open, free, unlimited.

kairix-agentic-knowledge-mgt

Shared knowledge layer for human-agent teams. Agents search, classify, and manage knowledge alongside your team.

mind-mem

Persistent AI memory for Claude Code, OpenClaw, and any MCP-compatible agent. BM25F + vector hybrid, governance-aware, local-first, zero-infrastructure.

inferedge-moss

Python SDK for semantic search with on-device AI capabilities

lodedb

World's fastest and most compact embedded vector database: exact by default, multimodal, local-first, and GPU-accelerated

vaultspec-rag

GPU-accelerated semantic search for your docs and source code - hybrid dense + sparse RAG on a local Qdrant backend, served to Claude Code and other MCP clients. The search companion to vaultspec-core.

knowledge-rag

Local RAG MCP server for Claude Code — hybrid search (semantic + BM25), cross-encoder reranking, 13 MCP tools, 20 format parsers. Zero external servers, zero API keys.

moss

The retrieval layer for production AI systems. Lightning-fast (<10ms) search without vector databases. Built for browser, edge, on-device, and cloud.

axon-rag

Local-first RAG engine with cited answers from your own docs. GraphRAG, sealed sharing, MCP-native, VS Code Copilot integration. Nothing leaves your machine.

dikw-core

AI-native knowledge base engine (client/server) across the DIKW pyramid: Data → Information → Knowledge → Wisdom. A self-managed server ingests, synthesizes and indexes; knowledge lives on disk as an open, portable Markdown format.

pyreindexer

Embeddable, in-memory, document-oriented database with a high-level Query builder interface.

trelix

🌲 Fast, reliable code intelligence — Tree-sitter AST parsing, contextual hybrid search, adaptive query planning, call-graph expansion & LLM synthesis. 20+ languages. Zero infra.

graph-tool-call

Graph-based tool retrieval for LLM agents — 248 tools → 82% accuracy, 79% fewer tokens. Zero dependencies. OpenAPI / MCP / LangChain.

overgraph

Absurdly fast embedded graph database with built-in vector search. Rust core, Node.js and Python connectors, GQL, graph algorithms, all in-process with no server.

mindcore-memory

Production-hardened MCP memory server for AI agents: hybrid BM25+FAISS search, circuit breaker, SLO tracking, Prometheus metrics, BND boundary evaluation. Gives AI persistent long-term memory that survives across sessions.

cocosearch

Local-first hybrid semantic code search tool. Indexes codebases into PostgreSQL with pgvector embeddings via Ollama, combines vector similarity + keyword search with RRF fusion. Supports 30+ languages. Features CLI, MCP server, WEB dashboard and interactive REPL.

trelix-mcp

🌲 Fast, reliable code intelligence — Tree-sitter AST parsing, contextual hybrid search, adaptive query planning, call-graph expansion & LLM synthesis. 20+ languages. Zero infra.

codeloom

Your AI agent's codebase map. Local-first, hybrid-search code graph for 51+ languages & formats. MCP server. 100% private.

kb-arena

Benchmark 9 retrieval architectures (vector, contextual, QnA, knowledge graph, hybrid, RAPTOR, PageIndex, BM25, rerank) on your own docs. Automated hyperparameter search with bootstrap CIs and significance tests.

trw-memory

Standalone AI agent memory engine — hybrid search (BM25 + vectors), Q-learning scoring, Ebbinghaus decay, tiered storage, knowledge graph. Part of TRW Framework.

faro-embedded-search

Incremental hybrid retrieval (lexical + vector, RRF-fused) with identical ranking on server Postgres and on-device SQLite