Consensus
AI-powered academic search engine. Ask a question, get evidence-based answers from 200 million peer-reviewed papers with a Consensus Meter showing scientific agreement.
What should journalists know about Consensus?
Consensus and Elicit occupy similar territory — AI-powered academic search — but take different approaches. Elicit is the researcher's tool: deep extraction, data tables, systematic review support. Consensus is the reporter's tool: ask a plain-language question, get a synthesized answer with citations and a visual meter showing how much the research agrees or disagrees. That Consensus Meter is the standout feature. When a politician claims 'studies show X,' you can type the claim into Consensus and see whether the weight of evidence actually supports it, partially supports it, or contradicts it. For fact-checking and science reporting, that's immediately useful. Consensus covers 200M+ papers — larger than Elicit's 138M — and returns results grounded in real research with direct citations. The Deep Search mode reviews up to 50 papers per query for thorough analysis. The company raised $19.2M including an $11.5M Series A from Union Square Ventures, Nat Friedman (former GitHub CEO), and Daniel Gross. Starting in the 2025-26 academic year, Consensus integrates with LibKey, letting university-affiliated users access paywalled articles through their library subscriptions. The limits: like any AI synthesis tool, it can oversimplify nuanced findings. The Consensus Meter works well for well-studied topics with large bodies of research but can be misleading for topics with only a handful of papers. Always read the actual studies it cites. The free tier is generous enough for most journalism use — 25 Pro searches and 3 Deep searches per month covers typical beat reporting needs.
Fact-checking scientific claims. Quick evidence synthesis on health, environmental, and policy topics. Assessing degree of scientific agreement on contested claims. Background research for science, health, and policy reporting. Verifying whether 'studies show' claims are real.
Deep systematic literature reviews (use Elicit for that). Topics with very few published studies — the Consensus Meter needs a body of research to be meaningful. Replacing your own reading of primary sources. Non-academic research questions (use general-purpose AI tools or web search).
Security & Privacy
Data is scrambled while being sent to their servers
Data is scrambled when stored on their servers
Where servers are located — affects which governments can request your data
Privacy policy summary
Consensus collects standard account data and search queries. Your queries reveal what scientific topics and claims you're investigating. The company is VC-backed with standard startup data practices. No published transparency report or SOC 2 certification. LibKey integration for university users routes through institutional library systems. Specific data retention and sharing policies require reviewing their full privacy policy.
How to protect yourself:
Your search queries reveal what scientific claims and topics you're investigating. For sensitive health or policy stories, consider whether query patterns could reveal an unpublished investigation angle. The free tier is generous enough that you can avoid creating an account for casual use. Verify every citation by reading the actual paper — AI synthesis can oversimplify nuanced findings. Don't treat the Consensus Meter as definitive for topics with few studies.
HTTPS encryption in transit. U.S. jurisdiction. VC-backed startup with standard security practices. No published SOC 2 certification or independent security audit. Search queries reveal your research interests, which is the primary privacy consideration for journalists. Adequate for academic background research and fact-checking. Be mindful that query patterns could reveal story angles for sensitive investigations.
Who Owns This
Pricing
Free: 25 Pro searches/month (20 papers each), 3 Deep searches/month (50 papers each), 10 GPT-4 Pro Analyses/month, 10 Study Snapshots/month. Pro: $15/month ($120/year) — unlimited core research features, monthly allotment of Deep searches. Enterprise and institutional plans available at custom pricing.
This is an editorial assessment based on publicly available information as of 2026-04-11, using our published methodology. Independent security review is pending. Security posture can change at any time. This is not a guarantee of safety.
Something wrong or outdated? Report it.