What is Elicit?

Elicit is an AI research assistant built specifically for working with academic literature. Instead of manually trawling databases, researchers type a question and Elicit searches across a corpus of more than 125 million papers, surfaces the most relevant ones, and pulls out structured data such as methods, sample sizes, and outcomes into a table you can sort and export. It also summarizes individual papers and can draft a systematic-review-style synthesis across a set of results, which is the part of literature review work that normally eats the most hours.

Who it is for

Elicit is aimed at academic researchers, PhD students, and R&D teams in industry who need to screen literature quickly, especially for systematic reviews, meta-analyses, or grant and thesis background sections. It is less useful for casual fact-checking or general web questions, where a broader tool like Perplexity or You.com fits better, and its coverage skews toward biomedical, life sciences, and social science fields rather than every academic discipline equally.

Key features

Semantic search across 125M+ papers, ranked by relevance to a natural-language question rather than keyword match
Automatic extraction of study data (population, intervention, outcomes, sample size, etc.) into a shared, exportable table
Per-paper summaries that highlight the key claim and methodology
Systematic-review style synthesis reports across a filtered set of papers
Chat interface to ask follow-up questions about a specific paper or the whole result set
Filters for study type, publication date, and citation count to narrow noisy searches

Pricing

Elicit offers a free tier with limited monthly credits for searches and extractions, and a Plus plan starting around $12/month that raises those limits and unlocks larger export tables. Higher-volume plans for labs and teams are available on request. Check elicit.com for current tier limits, since AI product pricing and credit allowances change often.

Strengths

Purpose-built for literature reviews rather than a general chatbot bolted onto a search box, so the workflow (search → extract → table → synthesize) matches how researchers actually work
Structured data extraction into tables is a genuine time-saver compared to reading and manually logging findings from dozens of papers
The Plus plan is inexpensive relative to the hours it can save on a review, and the free tier is enough to evaluate fit before paying

Weaknesses

Coverage has gaps outside biomedical and social science fields, so results can feel thin for niche engineering or humanities topics
Extracted data and summaries still need to be checked against the source paper — Elicit can misread tables or nuance in a methods section, and treating its output as final without verification is a real risk in a systematic review context

Getting started

Sign up at elicit.com with an email or Google account, then type a research question in plain language. Elicit returns a ranked list of papers; you can add columns to the table to extract specific fields, select papers to include, and generate a summary report once you have a filtered set you trust.

Verdict

Elicit is one of the more focused tools in AI Research & Search — it doesn't try to be a general-purpose answer engine like Perplexity, it tries to compress the literature-screening phase of research into hours instead of weeks. If your work involves systematic reviews or heavy academic reading, it's worth the Plus subscription; if you just need quick factual answers with citations, a broader research assistant will serve you better.