Consensus is built on a corpus of over 200 million peer-reviewed papers indexed from sources like PubMed, Semantic Scholar, and other academic databases. It uses large language models (powered by OpenAI's GPT infrastructure) to extract claims, synthesize findings, and generate consensus meters that indicate the degree of scientific agreement on a topic. The platform provides a Deep Filter system for narrowing results by study design, sample size, journal quality, and more. An API is available for developers to programmatically query the scientific literature corpus and integrate evidence-based answers into their own applications.