Skip to content

Ai2 OpenScholar

Answer science questions from retrieved papers

Visit official site

What Ai2 OpenScholar is

The model is described as a retrieval augmented language model built with the Allen Institute for AI and the University of Washington. It finds relevant papers and writes answers grounded in that literature across several scientific fields.

What you can do with it

  • Enter a science question in the demo search box
  • Read an answer built from retrieved papers
  • Check the accuracy and completeness of citations

Application scenarios

  • Biomedical staff ask about new drug research
  • Computer scientists summarise recent AI progress
  • Physicists collect literature on quantum computing

Pros & cons

✓ What we like

  • Iterative self-feedback refines the output
  • Evaluated by 20 scientists across fields

! What to watch out for

  • Answers depend on the papers that are retrieved
  • Users are told to check accuracy and citation completeness

FAQ

Who developed it?

The Allen Institute for AI and the University of Washington are named.

What is open-sourced?

All code, models, retrieval indexes and data are listed as open.

Last reviewed: 2026-11-15

More AI search engine tools

View all →

How we review