Elicit is an AI research assistant built specifically for scientific evidence synthesis. The product lets researchers ask a question in natural language, then searches across more than 200 million scholarly papers, extracts key data from full texts, summarizes findings, and surfaces contradictions or gaps in the literature.
Elicit is used by more than 400,000 researchers each month, including teams at Harvard, MIT, and the World Health Organization. The product is designed to make systematic reviews, literature reviews, and meta-analyses faster and more rigorous, with citation-grounded answers that are auditable back to source papers.
The company was originally incubated within Ought, a non-profit research lab focused on aligned AI. Elicit was founded in 2021 and is led by co-founder and CEO Andreas Stuhlmüller, who holds a PhD from MIT and conducted postdoctoral research at Stanford on automated reasoning. Co-founder Jungwon Byun has spoken publicly about the team's approach to supervising AI research workflows.
In February 2025 Elicit announced a $22 million Series A at a reported $100 million valuation, led by Spark Capital and Footwork, with existing investors Fifty Years, Basis Set Ventures, and Mythos Ventures participating. The capital is being used to expand the engineering team, deepen the data extraction stack, and broaden coverage of paper types and disciplines.
Elicit's differentiator is its focus on rigor: rather than generating answers from a generic chatbot, the product is engineered around process supervision and source-grounded reasoning, which is especially important in regulated and high-stakes domains like medicine, public health, and policy.