Consensus

About Consensus

AI that searches 200M+ scientific papers to find consensus on research questions. Get evidence-based answers with direct paper citations.

Consensus is a search engine for published research rather than the open web. You ask a research question in plain language and it searches a large index of scientific papers — it describes this as 200M+ papers — then summarises what the studies found, with direct citations to each paper so you can open the source. Its consensus meter gives a quick read on which way the published studies lean on a yes/no question.

That makes it a different tool from a web research agent. Our comparison of AI research agents puts it next to Elicit: Consensus gives a quick read on which way the evidence leans, while Elicit goes further into data extraction tables and systematic reviews. For peer-reviewed sources, both search papers rather than the open web, which keeps marketing pages and opinion pieces out of the answer.

How to use it well: treat the summary as a map, not a verdict. A meter that leans one way can still rest on small or low-quality studies, and the most common failure of any AI research tool is a real source that does not say what the summary claims. Open the cited paper and find the sentence before you repeat a finding, and export the citations into your reference manager when you write it up.

Pricing as listed here is freemium, with paid plans from $9 a month; check Consensus's own pricing page for the current limits. It suits healthcare, science and policy research, students and anyone who needs to know what the literature says rather than what the web says.