arXiv Search Scraper - Papers, Abstracts, PDFs for AI Agents

Use it to monitor new arXiv papers in a field, run literature reviews, or feed paper abstracts and PDF links to AI agents and RAG pipelines.

arXiv Scraper searches arXiv preprints by keyword, title, author, abstract or category and exports each paper with its title, abstract, authors, primary and secondary categories, submission and update dates, DOI, journal reference and PDF link. Use it to track new AI and machine learning papers, build research datasets, monitor authors or feed RAG pipelines with the latest preprints.

Why this Actor

• Official source. It uses the arXiv API at export.arxiv.org, with no HTML scraping.
• Respects arXiv limits. It sends one request every 3 seconds, as the arXiv API terms ask.
• Full query syntax. Use ti, au, abs and cat prefixes with AND, OR and ANDNOT.
• Use cases: daily new-paper monitoring in cs.AI or cs.CL, literature reviews, author tracking, trend analysis and RAG over preprints.

Example run

Real input and the first output record from run PwVNMZjRzmS4roAcH on 2026-10-30. Long text fields are shortened here with "...".

Input:

{
  "query": "all:\"retrieval augmented generation\"",
  "maxResults": 50,
  "sortBy": "submittedDate",
  "sortOrder": "descending"
}

Output (first record):

{
  "arxivId": "2610.40454v1",
  "title": "Glance: The Generator Is Its Own Reranker",
  "summary": "Retrieval-augmented generation commonly serves separate models for reranking and answering, duplicating model residency and passage processing. We introduce Glance, which equips the generator with a lightweight decision adapter and pointer ...",
  "authors": [
    "Adilet Metinov",
    "Chingiz Arziev"
  ],
  "published": "2026-10-29T17:01:22Z",
  "updated": "2026-10-29T17:01:22Z",
  "primaryCategory": "cs.CL",
  "categories": [
    "cs.CL",
    "cs.IR"
  ],
  "doi": null,
  "journalRef": null,
  "comment": "12 pages, 2 figures, 6 tables. Adapters and code: https://huggingface.co/airun-ai/GlanceRAG-MiMo-9B-Reranker",
  "pdfUrl": "https://arxiv.org/pdf/2610.40454v1",
  "url": "http://arxiv.org/abs/2610.40454v1"
}

Input

Field	Description
query	arXiv query, e.g. cat:cs.CL AND ti:transformer or au:hinton
maxResults	1 to 2,000
sortBy	submittedDate, lastUpdatedDate or relevance
sortOrder	descending or ascending

Output (one item per paper)

{"arxivId":"2401.00001v1","title":"...","summary":"...","authors":["Jane Doe"],"published":"2024-01-01T00:00:00Z","updated":"2024-01-02T00:00:00Z","primaryCategory":"cs.AI","categories":["cs.AI","cs.CL"],"doi":null,"journalRef":null,"comment":"12 pages","pdfUrl":"http://arxiv.org/pdf/2401.00001v1","url":"http://arxiv.org/abs/2401.00001v1"}

Notes

Thank you to arXiv for use of its open access interoperability. arXiv metadata is free to reuse; the papers themselves keep the license chosen by their authors. This Actor is not affiliated with or endorsed by arXiv.

Pricing

Pay-per-event: a small start fee plus a fee per paper returned.