LibRAG

AI Research project

LibRAG

AI Research builds LibRAG for source-grounded search, summaries and conversations across large cultural collections.

RAG for cultural heritage

Hybrid search, source citations, multilingual access The National Library of Kazakhstan provides partner collections; Yandex supports the project with cloud-service grant resources.

Author personas

Talk with authors represented in LibRAG

Choose an AI reconstruction grounded in the LibRAG collection and cached biography sources.

Books

2133 In the collection

Pages

465939 Indexed for search

Authors

936 Represented

Genres

14 Categories
Search

Browse the Collection

Explore books by genre, author, language, and year.

Browse Books
Chat

Chat with sources and authors

Ask about books, compare sources, or start a clearly marked AI reconstruction of an author.

Start Chat
Free mode

GreedSum summaries

Use extractive summarization without LLM chat for lightweight document review.

Compare plans

Choose how your institution works with LibRAG

Start with free access, enable LLM answers for a team, or discuss enterprise deployment by AI Research.