Consensus

Consensus is a search engine that runs on a database of more than 200 million academic papers instead of the open web. Type a question and it returns a list of matching papers with extracted findings, plus a “Consensus Meter” that gives a rough sense of whether the research leans for, against, or mixed on the question. It’s built to answer things like “does X actually work” without making you read ten abstracts to find out.

It suits students, researchers, journalists and anyone who wants to check a claim against actual published findings rather than a summarized blog post repeating someone else’s summary. The extracted-claims view saves real time over opening each paper individually, and the consensus scoring is useful for spotting when a topic looks more settled in popular writing than it is in the literature.

The catch is coverage: it leans heavily biomedical and life-sciences, so questions in economics, humanities or newer subfields turn up thinner results. It’s also a tool for finding and summarizing papers, not for judging whether any given paper is well-designed — a small, poorly controlled study and a large randomized trial can both show up as “supporting” evidence with equal visual weight, so the summaries are a starting point, not a substitute for reading the actual methodology.