Chunkr — нарезка документов для RAG и LLM

★ 4.1K

Chunkr is an open-source document-intelligence API: it turns PDFs, presentations, Word documents, and images into RAG/LLM-ready chunks via layout analysis, OCR, and semantic chunking. Reach for it when your RAG needs not just text from documents but well-cut, structured pieces: layout analysis, OCR with bounding boxes, structured HTML and Markdown output, and vision-language-model processing — yielding chunks ready for indexing. It deploys as a production-ready service. IMPORTANT: the open-source AGPL version differs from the fully managed Cloud API — the open-source release uses community models, while the Cloud API runs proprietary in-house models for higher accuracy and speed. Its focus is a service that turns documents into RAG-ready chunks via an API, not a parser library (docling) or a web app with a UI: the value is a ready pipeline for parsing and chunking documents for search and LLMs.

• #Documents