Hi hunters 👋 I built NextPaper because every time I found one good paper, the next hour went to manually chasing its references, its citations, and half-remembered searches — trying to figure out what else was worth reading. NextPaper does that in one click: open a paper (or type a topic), and it returns the most related papers, ranked by actual embedding similarity (SPECTER2) — not keyword matching — grouped into subtopics, with a timeline of the field, study design/sample size pulled from the abstract when it's unambiguous, and one-click citations in 9 styles (APA, MLA, Chicago, BibTeX...). A few things I care about that shape how it works: No server, no LLM, no account. Everything runs locally in your browser. Data comes from the free Semantic Scholar API — you bring your own free key, nothing is bundled or shared. Precision over coverage. If it's not sure a paper is a review, or what the sample size was, it shows nothing rather than guessing. A missing chip beats a wrong one. It's free and open source (MIT). I'm a solo dev/researcher and built this scratching my own itch during literature reviews, so I'd love to hear from people who actually do this work: what's missing, what's confusing, what would make you trust the rankings less or more. I'll be in the comments all day.

NextPaper

The related-papers tool that runs entirely in your browser

The related-papers tool that runs entirely in your browser