Elicit is an AI-powered research assistant built by Elicit, Inc. It automates literature reviews, evidence synthesis, and structured data extraction for scientific research. The platform queries over 138 million academic papers and 545,000 clinical trials using semantic search and natural language processing.
Users submit a natural language research question. Elicit retrieves relevant papers and screens them against custom inclusion and exclusion criteria. Extracted data populates interactive, customizable tables. A single session can analyze up to 1,000 papers and 20,000 data points. Enterprise plans extend paper screening to 40,000 papers.
Every AI-generated claim links to a sentence-level citation pointing to the exact quote or figure in the source paper. This design separates Elicit from general-purpose AI chatbots that produce unverifiable output. Independent benchmarks report 96% data extraction accuracy and 99.5% screening recall against 994 Cochrane reviews. A VDI/VDE policy review involving 1,511 data points recorded 99.4% extraction accuracy.
Over 5 million researchers use the platform. Customers include Oxford PharmaGenesis, Formation Bio, CSIRO, and Ultragenyx Pharmaceuticals. Users report up to 80% time savings on systematic reviews. Formation Bio extracted data 10 times faster than with manual methods. The tool is SOC 2 Type II certified.
Pricing
Elicit offers a free Basic plan with unlimited searches and paper summaries. Free users are limited to 2 automated reports and 2 table columns per month. The Pro plan costs $49 per user per month, billed annually at $588. It supports screening up to 5,000 papers with up to 20 table columns per report. The Scale plan costs $169 per user per month, billed annually at $2,028. It adds figure extraction, real-time collaboration, 200 data sources per report, and 30 table columns. Enterprise pricing is custom and extends screening to 40,000 papers with up to 40 table columns.
* Disclaimer: Please note that pricing information may not be up to date. For the most accurate and current pricing details, refer to the official website.
Key Features
-
✓
Semantic search across 138 million papers and 545,000 clinical trials
-
✓
PRISMA 2020-compliant systematic literature review workflows
-
✓
Large-scale data extraction into tables with up to 20,000 cells
-
✓
Sentence-level citations linking every AI claim to source text
-
✓
AI-generated synthesis reports of 10 or more pages
Use Cases
Systematic Literature Reviews
Research teams can automate end-to-end PRISMA-compliant literature reviews. Elicit handles protocol definition, database searching, paper screening against custom criteria, and data extraction into structured tables, supporting screening of up to 40,000 papers on Enterprise plans.
Clinical Trial Benchmarking
Pharmaceutical and medical teams can search the 545,000 integrated clinical trials to benchmark new trial designs. Users extract metrics by trial arm to identify endpoints, biomarkers, and predictive markers from past studies.
Regulatory Compliance Documentation
Medtech and pharmaceutical companies can generate auditable systematic reviews for regulatory filings such as IVDR and MDR. Every search and screening step is documented and traceable back to source data.
Evidence-Based Market Claims
Medical affairs and marketing teams can pull empirical data to substantiate product claims or compare a product against standard care. Sentence-level citations ensure that any supporting materials reference credible, verifiable research.
Exploratory Scientific Topic Research
Students and researchers entering an unfamiliar field can generate a structured AI Report covering the current state of a topic. The report summarizes findings from screened papers without requiring users to read dozens of individual studies from scratch.
Strengths & Weaknesses
Strengths
Every AI-generated claim links to a sentence-level citation in the source text, preventing unverifiable output.
Can analyze up to 1,000 papers and 20,000 data points in a single session.
Independent benchmarks show 96% data extraction accuracy and 99.5% screening recall.
Supports PRISMA 2020-compliant workflows with fully auditable search and screening records.
The built-in database of 138 million papers updates continuously, reducing reliance on manual uploads.
Weaknesses
The Free plan limits users to 2 table columns and 2 automated reports per month, making comprehensive research impractical.
Extracting data from figures and visual tables is restricted to the Scale plan at $169 per user per month.
Real-time team collaboration is only available on the Scale plan, not the Free or Pro tiers.
Accessing full text of paywalled papers requires institutional subscriptions or manual PDF uploads; Elicit does not bypass paywalls.
Who Is This For?
Pharmaceutical Researchers: Teams in drug discovery who need to screen large volumes of historical trial data, identify drug targets, and benchmark new clinical trial designs against existing evidence.
Medical Device (Medtech) Teams: Organizations that must produce PRISMA-compliant, auditable systematic literature reviews to satisfy regulatory requirements such as IVDR and MDR.
Academic Researchers and Students: Individuals conducting literature reviews for publications or dissertations who need to screen papers, manage citations, and remove duplicate references without spending weeks on manual work.
Policy Makers and Government Analysts: Analysts who must synthesize broad bodies of scientific evidence to design data-backed decisions in public health, education, or environmental policy.
Frequently Asked Questions
Is Elicit free to use?
Yes. Elicit offers a free Basic plan with unlimited searches and paper summaries. It limits table data extraction to 2 columns and automated reports to 2 per month.
What academic databases does Elicit search?
Elicit searches its own database of over 138 million academic papers, updated weekly. It also covers over 545,000 clinical trials from ClinicalTrials.gov, updated in real time, and includes a built-in PubMed search.
Can I upload my own research documents?
Yes. The BYOData feature allows users to upload proprietary PDFs. The Elicit Library also accepts reference metadata imported from Zotero, Mendeley, and EndNote.
Does Elicit use my data to train its AI models?
No. Elicit does not train on customer data. For Enterprise accounts, data is fully redacted, and third-party AI providers operate with zero data retention policies.
How accurate is Elicit’s data extraction?
Independent evaluations report 96% data extraction accuracy and 99.5% screening recall against 994 Cochrane reviews. A VDI/VDE real-world policy review of 1,511 data points recorded 99.4% accuracy.
How does Elicit prevent hallucinated citations?
Every AI-generated claim links to a sentence-level citation pointing to the exact quote or figure in the original source paper. Users can click through to verify the underlying text directly.
Which plan includes real-time team collaboration?
Real-time collaboration is available exclusively on the Scale plan ($169 per user per month) and Enterprise plans. It is not included in the Free or Pro tiers.
What are the paper screening limits per plan?
The Pro plan screens up to 5,000 papers per review. The Scale plan supports up to 200 data sources per report. The Enterprise plan extends screening to 40,000 papers and allows up to 40 table columns.
Does Elicit support regulatory compliance workflows?
Yes. Elicit supports PRISMA 2020-compliant systematic review workflows and is SOC 2 Type II certified. Its auditable search and extraction records are accepted in regulatory filings such as IVDR and MDR.
Can Elicit extract data from figures and charts inside papers?
Figure and visual table extraction is available on the Scale plan and higher. It is not included in the Free or Pro plans.
Elicit integrates with Zotero, EndNote, and Mendeley for importing reference libraries and metadata into the Elicit Library. PubMed and ClinicalTrials.gov are built-in search sources within the platform. Enterprise accounts support SSO and SAML for secure user authentication and account management. No third-party project management or productivity tool integrations are publicly documented.