Elicit: Features, Pricing, Models & Alternatives (2026)

Last researched: September 26, 2026. Plan prices and inclusions were checked against Elicit's own pricing page (https://elicit.com/pricing) and should be reconfirmed there, as Elicit has changed its tier structure more than once since launch. Elicit is built by Elicit, Inc., a public benefit corporation that spun out in 2023 from Ought, the nonprofit AI research lab that originally created it; Ought retains a stake in the resulting company.

What Is Elicit?

Elicit is an AI-powered research assistant purpose-built for literature review rather than open-ended chat. Instead of producing a single narrative answer, it finds relevant papers for a research question, extracts specific data points into sortable, structured tables, and cites the exact source sentence behind each claim it surfaces — a workflow aimed specifically at researchers doing systematic reviews, meta-analyses, or evidence synthesis.

Key Capabilities

• Research Agent automates multi-step literature search and synthesis for a given research question.
• Research Reports and Systematic Literature Reviews automate a substantial portion of the traditional manual review process.
• Structured data extraction into tables, with sentence-level citations for every extracted claim, rather than unsourced prose summaries.
• Screening support for reviewing up to thousands of papers against inclusion/exclusion criteria (scaling with plan tier).
• Custom extraction columns let researchers define exactly which data points to pull from each paper.
• Zotero import for bringing existing reference libraries directly into Elicit.

Main Features

Use Cases

• Conducting a systematic literature review or meta-analysis prep faster than fully manual screening
• Extracting comparable data points (sample size, methodology, outcome) across dozens or hundreds of papers into one table
• Getting sourced, citation-backed answers to a specific research question instead of an unsourced AI summary
• Setting up recurring research alerts to track new papers on a topic
• Importing an existing Zotero library to search and extract from within Elicit

Supported Models / Technologies

Elicit does not publish a single named "core model." It combines multiple large language models with its own extraction, ranking, and screening pipelines layered on top of the Semantic Scholar academic-paper corpus and ClinicalTrials.gov data. Historical reporting has associated Elicit with OpenAI's GPT-series models for portions of its pipeline, but current exact model versions and providers per feature are not publicly disclosed as of this writing.

Platforms

• Web (browser-based application only)
• API (Pro plan and above; unlimited API access on Enterprise)

No dedicated native desktop or mobile apps were identified as of this writing.

Integrations

Zotero (reference-library import), Semantic Scholar (underlying paper corpus), ClinicalTrials.gov (clinical trial data source), and API-based integrations for Pro/Scale/Enterprise customers.

Pricing

Advantages

• Structured, sourced data extraction is better suited to formal literature review than a general chatbot's prose summaries
• Generous free tier includes unlimited search across its full paper corpus
• Enterprise tier explicitly addresses research-integrity needs (no default data training, SSO/SAML, PRISMA-grade accuracy claim)
• Direct Zotero import reduces friction for researchers with existing reference libraries

Limitations

• Web-only; no native desktop or mobile apps
• Report/review generation is capped by plan even on paid tiers (e.g., a fixed number of reports per month on Pro)
• As with any AI extraction tool, results still require human spot-checking before being relied on in a formal published review
• Exact underlying model(s) are not publicly disclosed, which can matter for institutions with model-specific compliance requirements

Security & Privacy

Elicit states its Enterprise tier does not train on customer data by default and supports SSO, SAML, and two-factor authentication; for Basic, Pro, and Scale tiers, consult Elicit's own privacy policy and terms of service directly for the current data-training and retention policy, as those details were not independently itemized for every tier as of this writing.

Best For

Elicit fits academic researchers, graduate students, and evidence-synthesis teams (e.g., in medicine or the social sciences) who need to screen and extract structured data from many papers at once, rather than casual users looking for a single conversational research answer.

Alternatives

• Consensus - AI search engine focused on yes/no scientific-agreement questions rather than table-based extraction
• Semantic Scholar - the free, non-AI-synthesis academic search engine that underlies much of Elicit's own paper corpus
• Scite.ai - focuses on citation context ("smart citations") rather than data extraction
• ResearchRabbit - citation-graph-based literature discovery tool
• Litmaps - visual citation-mapping tool for tracking a research area over time

Frequently Asked Questions

Is Elicit free to use?

Yes. Elicit's Basic plan is free and includes limited usage of its Research Agent and Research Reports, plus unlimited search, summaries, and chat across its paper database, according to Elicit's own pricing page.

What database does Elicit search?

Elicit indexes more than 125-138 million academic papers (reporting varies by source and date) primarily via the Semantic Scholar corpus, plus roughly 545,000 clinical trials sourced from ClinicalTrials.gov.

Which AI models does Elicit use?

Elicit does not publish a single named core model; it combines multiple large language models with its own extraction and ranking pipelines built on the Semantic Scholar and ClinicalTrials.gov datasets. Exact current model versions and providers are not publicly disclosed as of this writing.

Is Elicit accurate enough for a formal systematic review?

Elicit markets its higher tiers toward systematic-review workflows and its Enterprise tier specifically claims "PRISMA-grade accuracy," but as with any AI research tool, extracted data and summaries should be spot-checked against source papers before being relied on in a published review.

Final Summary

Elicit's distinguishing choice is structured, citation-backed data extraction over conversational summarization, aimed squarely at literature-review workflows rather than general research chat. Its free tier is genuinely usable for casual search, but the report/review caps on paid tiers mean heavy systematic-review users should budget for Pro or Scale rather than assuming Basic will cover a full project.