

























litscout
“Art and curiosity shall not be killed by finitude”
AI is Researcher Artist
AI is a tool.
You imagine, AI helps.
You are the artist.
YOU
A research paper search engine that ranks by intent, not keywords
Most literature search matches words and sorts by citation count — which favours older work and assumes you already speak the field's vocabulary. litscout models what you are trying to find out, and ranks against that.
Start a systematic literature review
Describe the question rather than guessing the keywords. litscout extracts the requirements and constraints from what you wrote, then searches along several paths at once — exact terms, meaning, and the citation neighbourhood of anything you already know is relevant.
Follow the citation graph in both directions
Backward and forward snowballing from any seed paper: what it cites, what cites it, what sits beside it through co-citation and bibliographic coupling. This is where the papers keyword search never surfaces actually live.
Search by DOI, arXiv ID, ISBN or a pasted abstract
References are high-confidence signals, so they are treated as such. Paste a DOI, an arXiv ID, a patent number, a half-remembered title or an entire abstract, and it resolves to a canonical record and searches outward from there.
Find preprints, patents, theses and datasets
Not only journal articles. Patents, standards, technical reports, theses, benchmark suites and datasets are first-class and filterable — which matters for applied questions, where the answer is often a standard nobody indexed as a paper.
See why every result was returned
Each result carries the reason it was selected: the requirement it satisfies, how it relates to a paper you flagged, where it sits in the citation graph. A ranking you cannot interrogate is one you cannot rely on.
Export to Zotero, Mendeley or EndNote
Results and collections export to BibTeX and RIS, which every major reference manager imports directly. Nothing here is a walled garden.
Already using Google Scholar? Here is an honest comparison, including where Scholar is the better tool.