Epsilon

Epsilon is an AI search engine designed for scientific research, enabling users to search over a large academic corpus, generate summaries with inline citations, and synthesize insights across multiple papers. It provides tools to find publications and patents, extract information from multiple sources, and build organized libraries for literature reviews and meta-analyses.

Users enter research questions and Epsilon searches over a dataset of more than 200 million papers, returning the top 100 related papers. It generates a ChatGPT-like summary with inline citations to underlying sources. The platform allows saving papers to a library, organizing libraries, and running cross-library searches to synthesize results. It can extract information from multiple papers for meta-analyses, group results into latest research and most relevant articles, and provides access to publications and patents directly from results. The system leverages Semantic Scholar data and uses GPT-4 to produce summaries while maintaining citation-based trust.

Who it’s for: Researchers and academics conducting literature reviews, meta-analyses, grant writing, and evidence synthesis across scientific disciplines

Website shows product shutdown date (April 30, 2026), indicating nearing end-of-life; mentions adoption by researchers and ongoing blog/updates, suggesting active user traction prior to shutdown