Academic Paper Search – Crossref, arXiv & PubMed (Europe PMC)

Search more than 150 million scholarly papers across Crossref, arXiv and Europe PMC (PubMed + PubMed Central) in one run — and get clean, deduplicated JSON with titles, authors, abstracts, DOIs, citation counts, journals and open-access PDF links. A fast, reliable Google Scholar alternative built only on official open scholarly APIs: no CAPTCHAs, no blocked runs, no grey-zone scraping.

• 📚 Three sources, one schema: Crossref (all disciplines, 150M+ DOIs, citation counts), arXiv (CS, AI, physics, maths, stats preprints with PDFs), Europe PMC (life sciences & medicine: PubMed, PMC, bioRxiv/medRxiv preprints)
• 🔎 Filters & sorting: year range, open access only, sort by relevance, newest or most cited
• 🧬 Rich metadata: authors with ORCID and affiliation, abstract, venue, publisher, volume/issue/pages, subjects/MeSH keywords, license, PMID/PMCID/arXiv ID
• 📄 Open-access PDF links where available (always for arXiv)
• 🔁 Deduplication across sources by DOI, arXiv ID or title + year
• 🔗 DOI lookup: paste DOIs to fetch full metadata for a reference list
• 💸 $1 per 1,000 papers — no subscription, no per-query fees

Who is it for?

• Researchers & students: literature reviews, systematic review screening exports, reading lists
• AI / RAG builders: feed abstracts and PDFs into LLM pipelines, build domain corpora, ground agents in real citations
• R&D, pharma & biotech: monitor new publications on a target, compound or method (schedule it weekly with sort: newest)
• Analysts & funders: track research trends, top-cited work, venues and institutions in a field
• Librarians & publishers: enrich reference lists by DOI

Pricing

Event	Price
Paper returned	$0.001 ($1 per 1,000 papers)

Example: 3 queries × 3 sources × 50 papers = up to 450 papers ≈ $0.45 (less after deduplication). Duplicates removed by dedup are not charged; failed DOI lookups are free.

For comparison, Google-Scholar-based Actors in the Apify Store charge roughly $0.0015–$0.008 per paper or $0.015 per query (October 2026) and depend on scraping a site that blocks bots. Set Maximum cost per run to cap spending.

Input example

{
  "queries": ["retrieval augmented generation evaluation", "\"large language models\" hallucination"],
  "sources": ["crossref", "arxiv", "europepmc"],
  "maxResultsPerSource": 50,
  "sort": "citations",
  "yearFrom": 2023,
  "openAccessOnly": false
}

DOI lookup:

{ "dois": ["10.1038/nature14539", "https://doi.org/10.1145/3641289"] }

Output example

{
  "query": "retrieval augmented generation evaluation",
  "status": "ok",
  "source": "arxiv",
  "sourceId": "2411.18583v1",
  "title": "Automated Literature Review Using NLP Techniques and LLM-Based Retrieval-Augmented Generation",
  "authors": [{ "name": "Nurshat Fateh Ali", "orcid": null, "affiliation": null }],
  "authorNames": "Nurshat Fateh Ali, Md. Mahdi Mohtasim, Shakil Mosharrof, T. Gopi Krishna",
  "authorCount": 4,
  "year": 2024,
  "publishedDate": "2024-11-27",
  "venue": "arXiv",
  "type": "preprint",
  "doi": null,
  "url": "https://arxiv.org/abs/2411.18583",
  "pdfUrl": "https://arxiv.org/pdf/2411.18583v1",
  "abstract": "This research presents and compares multiple approaches to automate the generation of literature reviews…",
  "citationCount": null,
  "isOpenAccess": true,
  "subjects": ["cs.CL", "cs.AI", "cs.IR", "cs.LG"],
  "arxivId": "2411.18583"
}

Crossref and Europe PMC rows add citationCount, venue (journal), publisher, volume/issue/pages, license, and pmid/pmcid where available.

Views: Papers (title, authors, year, venue, citations, links) and Abstracts. Export to CSV/Excel for Zotero, Mendeley, EndNote or spreadsheets, or use the API/JSON.

Tips

• Exact phrases: wrap in quotes — "graph neural networks".
• arXiv power syntax: ti:transformer AND cat:cs.CL, au:hinton.
• Europe PMC syntax works for the Europe PMC source, e.g. TITLE:"long covid" AND METHODS:cohort.
• Most cited uses each source's citation counts (Crossref "is-referenced-by", Europe PMC "cited by"); arXiv has no citation counts and is returned by relevance.
• Monitoring: schedule weekly with sort: "newest" and yearFrom = this year, and send results to email/Slack/Sheets via integrations.

Data sources and terms

All data comes from official, free scholarly metadata APIs used within their published rate limits and terms:

• Crossref REST API — bibliographic metadata deposited by publishers (facts and metadata; abstracts where publishers provide them)
• arXiv API — "Thank you to arXiv for use of its open access interoperability." Requests are throttled to arXiv's limit of one call every 3 seconds.
• Europe PMC RESTful API — PubMed/MEDLINE, PMC and preprint records

This Actor returns metadata and links; it does not download or redistribute full-text articles. Respect each paper's license when using full texts.

FAQ

Is this Google Scholar? No — Google Scholar has no API and forbids automated access. This Actor uses open scholarly databases that cover the vast majority of peer-reviewed literature and preprints, with stable, structured results.

Why do some papers lack abstracts? Not all publishers deposit abstracts in Crossref. Europe PMC and arXiv records almost always include them.

Why are citation counts different from Google Scholar? Each database counts citations from its own index; Crossref and Europe PMC counts are usually lower than Google Scholar's but consistent and reproducible.

How many results can I get? Up to 5,000 per source per query; Crossref and Europe PMC use cursor paging, so large exports are reliable.

Related Actors

• PDF to Markdown — convert open-access PDFs into clean Markdown for LLMs
• AI Web Scraper — extract structured fields (methods, datasets, results) from paper landing pages
• RSS Feed Reader & Monitor — follow journal RSS feeds for new issues