arXiv APIunofficial
Public arXiv paper abstracts, search results and category submissions from arxiv.org HTML pages.
- arxiv.org
all live 4 operations, checked by the installs that call them.
Seeded by the maintainers.
Try it
Set up API Anything for your agents:
npx -y api-anything setupThen ask your agent, which finds this site by itself, or call it from a shell:
api-anything pull arxiv
api-anything call arxiv getPaper id=1706.03762Operations
| Operation | Method | Login | Writes | Status |
|---|---|---|---|---|
getPaper(id) -> id, title, authors, abstract, submitted, pdfUrl, versionUrl, primarySubject, subjects, comments, journalRef, doiRead a paper abstract page and its bibliographic metadata. |
GET | no | no | Live0 confirmations · success unknown; last worked unknown |
listNew(category) -> id, title, authors, subjectsList the latest new submissions in an arXiv category. |
GET | no | no | Live0 confirmations · success unknown; last worked unknown |
listRecent(category) -> id, title, authors, subjectsList recent submissions in an arXiv category. |
GET | no | no | Live0 confirmations · success unknown; last worked unknown |
searchPapers(q, searchtype?, order?, start?) -> id, title, authors, abstract, submitted, subjectsSearch public papers by field, sort order and result offset. |
GET | no | no | Live0 confirmations · success unknown; last worked unknown |
Notes
No account is needed. Paper licenses vary; check the license on each abstract page before reuse.
getPaperaccepts modern IDs such as1706.03762and legacy IDs such ashep-th/9711200. It returns one item.versionUrlidentifies the version shown, usually the latest when no version is requested. Comments, journal reference and DOI can be absent. Authors use the citation metadata's surname-first format; subjects are page text.searchPapersdefaults to all fields and the first 50 results in announcement-date order. Usesearchtype=authorfor author search,orderfor sorting, andstart=50,100, etc. for subsequent pages. An emptyorderselects relevance. The abstract is a search snippet and may contain a “More” marker.listNewreturns the new-submissions section, excluding cross-lists and replacements. Its title and subject text retain the page's labels. The page changes on announcement days, and may limit long lists; api-anything may also truncate large responses.listRecentreturns the first page of the category's recent submissions as served by arXiv. Its title and subject text retain the page's labels; lists change over time and may be truncated.
About these operations
Each operation is a request learned from arXiv's own web pages: API Anything replays it over plain HTTP and falls back to a browser when it must. Read them as JSON.