Description
Documents full of tables, images and formulas defeat text-only RAG. RAG-Anything from HKU's HKUDS is an all-in-one multimodal RAG framework handling text, images, tables and equations together.
It integrates natively with LightRAG to bring multimodal RAG there too.
Multimodal parsing:Text, images, tables and formulas.
Unified retrieval:Cross-modal Q&A.
LightRAG:Native integration.
It integrates natively with LightRAG to bring multimodal RAG there too.
Features
Multimodal parsing:Text, images, tables and formulas.
Unified retrieval:Cross-modal Q&A.
LightRAG:Native integration.
Tags:rag