Consensus

Consensus is an AI search engine designed specifically for finding what the scientific literature says about a question. Rather than returning a list of papers to read, it analyses the available peer-reviewed evidence and presents a consensus summary indicating what direction the research points, along with the most relevant papers supporting that view.

The core feature is the Consensus Meter, which visualises the degree of agreement across studies on a given claim. Searching a question like 'does intermittent fasting improve metabolic health' returns not just relevant papers but a visual indicator of whether the available studies predominantly support, oppose, or are divided on the claim. This is useful for policy researchers, evidence-based practitioners, and non-specialists who need to understand the state of evidence on a topic without deep domain expertise.

The tool covers a large corpus of peer-reviewed papers and is updated regularly. The quality of synthesis is generally good for well-studied topics where a substantial body of evidence exists. For niche topics or emerging areas of research, the database coverage can be thin and the consensus view may not represent the actual state of knowledge.

Consensus is complementary to Elicit rather than a direct replacement. Where Elicit is more useful for systematic literature review with detailed data extraction, Consensus is better for quickly understanding the general direction of evidence on a question. Many researchers find both tools useful for different stages of the research process.