Elicit

Elicit is an AI platform for scientific research that searches over 138 million papers and 545,000 clinical trials, generates cited research reports, automates systematic-review screening and data extraction, and organizes sources through a Library with alerts for new publications.

Elicit is an AI-driven research platform built to help scientists, academics, and industry R&D teams work through published scientific literature faster and with more rigor. Its core is semantic search across more than 138 million academic papers and over 545,000 clinical trials, meaning researchers can surface relevant work without knowing the exact terminology original authors used. Beyond search, Elicit generates structured research reports with sentence-level citations, screens and extracts data from large paper sets, and lets users interrogate sources directly for evidence-based answers. The tool is built for anyone who needs to synthesize scientific evidence at scale, spanning academia, pharmaceuticals, medical devices, policy and government, consumer goods, industrials, and software teams. Documented case studies describe a pharmaceutical team cutting data-extraction time roughly tenfold across 1,600 papers, an independent reviewer reaching over 99% data-extraction accuracy on a policy-informing systematic review, and a research consultancy scaling reviews across 40 questions and 500 papers for pharma clients. In practice, a user starts with a research question rather than a keyword search. Elicit screens matching papers against inclusion criteria, extracts structured data points into tables, and compiles a report modeled on systematic-review methodology. Reports are customizable, so teams can control which papers are considered and which fields get extracted, and the Systematic Review workflow now supports PRISMA 2020 guidelines with traceability at each step. A Library stores and organizes sources for reuse across projects, and Alerts flag new relevant publications so long-running projects stay current without repeated manual searching. What separates Elicit from general-purpose AI chat tools is its emphasis on scale and verifiability: a single workflow can process up to 1,000 relevant papers and analyze up to 20,000 data points, with every generated claim traced back to a specific source sentence. An API and MCP server extend these capabilities into custom agents and pipelines, positioning Elicit less as a one-off search tool and more as infrastructure for evidence-based research, sitting alongside general AI assistants and dedicated systematic-review software as an alternative built specifically for scientific literature work.