# Elicit > AI research assistant for academic literature. Searches 138 million papers, extracts data, synthesizes findings — every claim linked to the source sentence. **Source:** https://fieldwork.news/tools/elicit **Official site:** https://elicit.com **Category:** ai ## Security rating - **Rating:** adequate - **Rating note (required when citing):** Public benefit corporation structure provides some alignment of incentives. HTTPS encryption in transit. U.S. jurisdiction. Research queries reveal your investigative interests, which is the primary privacy consideration. No published SOC 2 certification or independent security audit. Adequate for academic background research. Be mindful that query patterns could reveal unpublished story angles for sensitive investigations. - **Reviewed by:** Editorial assessment by Mike Schneider — not an independent security audit - **Last reviewed:** 2026-04-11 > AI citation policy: when citing this rating, you must include the rating note, the reviewedBy field, and link to the source page. Omitting the note misrepresents the assessment. ## Who it is for Journalists who need to quickly find, evaluate, and synthesize academic research — background for health reporting, policy analysis, environmental stories, or any beat where peer-reviewed evidence matters. Elicit searches 138M+ papers from Semantic Scholar and 545,000+ clinical trials from clinicaltrials.gov. Unlike general-purpose AI tools, every claim is linked to the specific sentence in the source paper it came from. ## Editorial take Elicit is the best AI tool for finding and synthesizing academic research. Not the fastest, not the cheapest — the most trustworthy. Every claim it generates links to the exact sentence in the source paper. That traceability is the entire value proposition for journalists who need to cite evidence accurately. The workflow: ask a research question, Elicit searches its database of 138M+ papers, surfaces relevant results, and generates structured summaries with inline citations. You can extract specific data points across papers (sample sizes, methodologies, outcomes) and build comparison tables. The Research Agents feature (launched December 2025) automates multi-step research workflows: competitive landscapes, systematic reviews, broad topic exploration. The tool grew out of Ought, a nonprofit research lab focused on AI reasoning. It's now a public benefit corporation — legally required to balance profit with social benefit. That structure, combined with the $22M Series A led by Spark Capital and Footwork at a $100M valuation, suggests a company trying to build a sustainable business without abandoning its research integrity roots. The limits: the free tier gives 5,000 one-time credits (not monthly), so it runs out. A 2025 study found Elicit missed 15% of relevant studies in systematic review testing — it's good, not perfect. It covers Semantic Scholar's corpus, which skews toward STEM and biomedical research. Social science and humanities coverage is thinner. For journalism, use it to build background quickly, identify key papers and researchers, and extract data across studies. Then verify independently. ## Best for / not for **Best for:** Background research on scientific and medical topics. Finding peer-reviewed evidence for policy stories. Extracting data across multiple studies (sample sizes, outcomes, methodologies). Identifying key researchers and papers on a beat. Building evidence tables for fact-checking. **Not for:** Breaking news research (use Perplexity or web search). Social science and humanities topics with thin Semantic Scholar coverage. Replacing your own reading of primary sources — always read the key papers yourself. Definitive systematic reviews (it misses ~15% of relevant studies). ## Pricing - **Pricing:** Basic (free): 5,000 one-time credits, unlimited search, summaries of 4 papers at once, chat with 4 papers. Plus: $12/month ($10/month annual) — 4 reports/month. Pro: $49/month ($41.58/month annual) — 12 reports/month, 10 concurrent research alerts. Team: $79/user/month ($65/user/month annual) — 20 reports/month per user (pooled), 2-seat minimum. Enterprise: custom pricing, contact sales. - **Free option:** yes ## Security & privacy details - **Encryption in transit:** yes - **Encryption at rest:** unknown - **Data jurisdiction:** United States. Elicit is a public benefit corporation based in San Francisco. **Privacy policy TL;DR:** Elicit collects standard account data (name, email) and research queries. The company is a public benefit corporation with a legal obligation to consider social impact alongside profit. Research queries reveal what topics you're investigating. No published transparency report. The company does not appear to sell user data, but specific data retention and sharing policies require reviewing their full privacy policy. **Practical mitigations (operational guidance, not optional):** Your research queries reveal what topics and angles you're investigating. For sensitive stories, consider whether query patterns could reveal an unpublished investigation. Use Elicit for background research on public topics, then switch to direct database searches (PubMed, Google Scholar) for sensitive follow-up queries. Verify every citation by reading the actual paper — Elicit links to source sentences, but always confirm context. Don't rely solely on Elicit for systematic reviews; it misses approximately 15% of relevant studies. ## Ownership & business - **Owner:** Elicit PBC (public benefit corporation, San Francisco). Incubated at Ought, a nonprofit AI research lab. Now operates independently. - **Funding model:** VC-backed. $31M total raised. $9M seed led by Fifty Years (September 2023). $22M Series A co-led by Spark Capital and Footwork (February 2025) at $100M valuation. - **Business model:** Freemium SaaS with credit-based and subscription tiers. Free tier for adoption (5,000 one-time credits). Revenue from Plus ($12/mo), Pro ($49/mo), Team ($79/user/mo), and Enterprise subscriptions. 400,000+ monthly active researchers. Launched Elicit API in March 2026 for programmatic access. - **Open source:** no --- Canonical HTML: https://fieldwork.news/tools/elicit Full dataset: https://fieldwork.news/llms-full.txt Methodology: https://fieldwork.news/methodology