Libraries & frameworks
chonkie
chonkie-inc/chonkie
A lightweight Python chunking library for RAG pipelines with many chunkers, refiners, embeddings, and vector-database integrations.
Topics
Related projects
R2R8.1kA production-ready agentic RAG system with multimodal ingestion, hybrid search, knowledge graphs, and a REST API.
kreuzberg9.6kA polyglot document-intelligence engine that extracts text, tables, and metadata from 100+ formats via libraries, CLI, REST, and MCP server.
USearch4.4kUSearch is a compact, single-header C++ similarity search and clustering engine for vectors and arbitrary objects. It offers fast HNSW indexing, multiple language bindings, custom metrics, quantization, and disk-backed indexes for embedding it in applications.
spacy-layout914A spaCy plugin built on Docling that turns PDFs and Word files into structured Doc objects with layout spans, tables and Markdown for NLP and RAG pipelines.
ragflow93.8kAn open-source RAG engine that gives LLMs grounded retrieval over complex documents.
mem068.8kA memory layer SDK for AI agents and assistants that stores, retrieves, and personalizes long-term context.