Elicit

Research assistant that extracts findings from papers into a table

AI Search & KnowledgeFree planOverseasβ˜…β˜…β˜…β˜…β˜† 4.0

What is Elicit?

Elicit searches academic literature and pulls structured data out of the papers it finds, so a question becomes a table with columns such as population, method and outcome rather than a list of links. Every extracted cell cites the source it came from, and the workflow supports screening steps that resemble a systematic review. It is built by Ought and aimed at evidence synthesis: comparing many studies along one axis is the task it is designed for. It is a hosted commercial product with a limited free allowance, and the extraction step still needs a human check before anything is relied on.

Last updated: 2026-09-20. This site only provides an index; for exact features, pricing, and licensing, see the official website.

Key features

  • Literature search that returns a summarised list of candidate papers
  • Automatic extraction of chosen fields into a comparison table
  • A source citation in each extracted cell so values can be traced
  • Screening and inclusion steps that mirror a review workflow
  • Custom columns defined by the user for a specific question
  • Export of tables and references for further analysis

Pros & cons

Strengths

  • Turns dozens of papers into a comparable table quickly
  • Every extracted value links back to the paper it came from
  • Workflow mirrors evidence synthesis rather than casual search

Watch out for

  • Extraction errors are common enough that cells need checking
  • The free allowance is small relative to a real review
  • Coverage and extraction quality vary noticeably by discipline

Best for & use cases

evidence synthesis, systematic review prep, comparing studies and data extraction

If you're comparing similar products, check the alternatives below, or browse all tools in the AI Search & Knowledge category.

FAQ

Can I trust the extracted values?

Treat them as a first pass. Elicit links each value to its source, and that link exists precisely because manual verification is expected, particularly for numbers, sample sizes and study design.

What kind of questions suit Elicit?

Comparative questions work best: which interventions were tested, what outcomes were measured, in which populations. Broad exploratory questions are better served by a citation map or a search engine.

Does it write the review for me?

No. It assembles and structures evidence. Interpretation, quality appraisal and the argument remain yours, which is also what makes the output defensible in a published review.