Elicit
AI research assistant that turns questions into structured extraction tables across 138M+ papers.
About
Elicit (by the nonprofit Ought) is built for systematic literature work: ask a research question in natural language and it returns a table of relevant papers with automatically extracted data points — sample sizes, methods, outcomes — which you define as columns. It also offers a Research Agent (Pro+) that searches beyond papers into trials and regulatory documents.
It is academic-only by design: no market research, no general web. Automated extraction can misread tables and nuance — treat it as a fast first pass, then verify against the papers before citing. Pricing: free Basic tier; paid Plus (~$12/mo), Pro (~$49/mo) and Scale tiers with annual discounts (rates vary by academic vs industry audience — check the pricing page).
What it can do
- Semantic search across 138M+ papers
- Structured extraction tables with custom columns
- Systematic-review screening workflows (Pro)
- Research Agent for trials/regulatory docs (Pro+)
- CSV, BibTeX, RIS export; API access (Pro+)
Setup
- Sign up at elicit.com (free Basic tier, no card).
- Enter a focused research question in the search bar.
- Review retrieved papers and one-sentence abstract summaries.
- Add extraction columns (population, intervention, findings) and let Elicit fill the table.
- Export to CSV, BibTeX, or Zotero.
Prerequisites
A web browser; Elicit account (free tier available).
Use cases
Works with
Trust score — 7/10
Scored 7/10 (metrics tier). Strong product with broad independent 2026 review coverage and a nonprofit operator, but closed SaaS (openness capped) and no public adoption metric beyond provider-tier estimates. Docked for extraction-accuracy caveats flagged by reviewers and pricing that varies by audience. Solid metrics-tier entry.
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 8.0/10
- Maintenance · recency, releases, issue load
- 5.0/10
- Openness · license + docs
- 4.8/10
- Manual review · sanity read + red-flag check
- 9.0/10
- · provider-tier estimate — thinner evidence (no public adoption metric)
- · widely used in academia; no public user metric — provider-tier estimate
- · no release date — neutral maintenance score
- · manual review: strong independent coverage (multiple 2026 reviews); pricing verified; extraction errors possible — verify against papers; license: closed SaaS (Ought nonprofit)
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.