Academic Paper Search – arXiv, PubMed, DOI

sfoln/academic-paper-search · AI, Developer tools, Automation On request

Search research papers across Crossref, arXiv, PubMed and Europe PMC in one call, or look up DOIs, arXiv ids and PMIDs. One merged, de-duplicated row per paper: title, authors, year, venue, DOI, abstract, citation count, open-access PDF link. $1 per 1,000 papers.

Request access Run it$1 / 1,000 research paper

About

What it does

Search research papers across Crossref, arXiv, PubMed and Europe PMC in one call and get one clean, de-duplicated row per paper: title, authors, year, publication date, journal or venue, DOI, arXiv id, PMID, abstract, citation count, open-access flag with the best free PDF link, and subjects. Or paste DOIs, arXiv ids, PMIDs or PMCIDs and get the same record for each. Built for literature reviews, research assistants and AI agents that need reliable paper metadata without scraping Google Scholar. $1 per 1,000 papers.

Price

Price per delivered record

Event	Price	Per 1,000	Charged for
Research paper `paper`	$0.001	$1 / 1,000	One paper delivered: a search result row (merged from Crossref, arXiv, PubMed and Europe PMC) or the record of a DOI, arXiv id, PMID or PMCID you looked up.

Apify free plan: up to 10 items per run. Paid plans have no per-run cap; set maxTotalChargeUsd to cap spend.

Run

Call it from code, an agent or the Console

This API is private on Apify for now. Email us and we will publish it for you, usually within two working days. The calls below work once it is public. contact@sfoln.com

curl · synchronous run, rows in the response

curl -X POST "https://api.apify.com/v2/actors/sfoln~academic-paper-search/run-sync-get-dataset-items" \
  -H "Authorization: Bearer $APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"queries":["crispr gene editing"],"maxResultsPerQuery":5}'

mcp.json · hosted Apify MCP server

{
  "mcpServers": {
    "apify": {
      "type": "http",
      "url": "https://mcp.apify.com?tools=sfoln/academic-paper-search"
    }
  }
}

asynchronous run · for large batches

POST https://api.apify.com/v2/actors/sfoln~academic-paper-search/runs?maxTotalChargeUsd=1
GET  https://api.apify.com/v2/actor-runs/{runId}            # until status is SUCCEEDED
GET  https://api.apify.com/v2/datasets/{defaultDatasetId}/items?format=json

Authenticate with an Apify API token. MCP clients can instead sign in with OAuth when prompted. Example prompt for an agent: Find the 5 most relevant papers on "crispr gene editing" published since 2022 and list their titles, authors, journals, DOIs and citation counts.

Input

Input fields

Search research papers across Crossref, arXiv, PubMed and Europe PMC, or look up papers by DOI, arXiv id, PMID or PMCID. Each paper is one dataset row; each query also gets one summary record.

Field	Type	Default	Meaning
`queries`	`array`	—	Search queries Topics or keywords to search, one query per line (e.g. "crispr gene editing"). Words must all appear; put a phrase in double quotes. A line that is a DOI, an arXiv id or URL, or "PMID: …" is looked up instead. Accounts on the Apify free plan can process up to 10 items per run (queries and identifiers together) and 20 papers per query.
`ids`	`array`	—	DOIs, arXiv ids, PMIDs or PMCIDs Papers to look up, one per line: DOIs (10.1038/s41586-021-03819-2 or a doi.org link), arXiv ids or links (1706.03762, arXiv:hep-th/9901001), PubMed ids (34968414) or PubMed Central ids (PMC8371605). Each returns one record.
`maxResultsPerQuery`	`integer`	`25`	Max papers per query How many papers each search query returns at most (merged across the sources, best first). Range 1–1000.
`sources`	`array`	—	Sources Where to search: crossref (all disciplines, citation counts), arxiv (physics, maths, computer science preprints), pubmed (biomedical literature, MeSH terms), europepmc (life sciences incl. preprints, open-access full text), openalex (only when available on this Actor). Leave empty for all of them.
`yearFrom`	`integer`	—	Published from (year) Only papers published in or after this year. Range 1000–2999.
`yearTo`	`integer`	—	Published until (year) Only papers published in or before this year. Range 1000–2999.
`type`	`string`	`"any"`	Publication type Only this kind of publication. Sources without such records are skipped (arXiv only has preprints). One of `"any"`, `"journal-article"`, `"preprint"`, `"review"`, `"conference-paper"`, `"book"`.
`openAccessOnly`	`boolean`	`false`	Open access only Only papers that are free to read (open-access licence, arXiv, PubMed Central, Europe PMC full text).
`sort`	`string`	`"relevance"`	Sort by relevance: papers several sources rank high come first; date: newest first; citations: most cited first. One of `"relevance"`, `"date"`, `"citations"`.
`enrich`	`boolean`	`true`	Add citation counts by DOI For papers with a DOI that Crossref did not return in the search, fetch the Crossref record (citation count, licence, publisher). No extra charge.
`deduplicate`	`boolean`	`true`	Skip papers already delivered in this run A paper found by several queries is delivered (and charged) once, for the first query.
`maxItems`	`integer`	`100`	Max items Safety cap on how many queries and identifiers are processed in this run; the rest are reported as ignored. Range 1–10000.
`concurrency`	`integer`	`3`	Concurrency Queries and lookups processed in parallel. arXiv and Crossref limits are shared by the whole run, so more rarely helps. Range 1–6.
`includeRaw`	`boolean`	`false`	Include raw source records Attach each source's unmodified record to every paper (larger output; with queries of 500+ papers give the run 1 GB of memory).

Example input

{
  "queries": [
    "crispr gene editing"
  ],
  "maxResultsPerQuery": 5
}

Output

One record per input item

Records land in the run's dataset; a synchronous call returns them directly. Each has a status (ok is charged; everything else is free) and a charged flag. The run's OUTPUT record totals items and charges.

Field	Type	Meaning
`input`	`string \| object`	Input The search query or identifier this record belongs to.
`recordType`	`string`	Record type paper (one paper; charged) or summary (one per search query: counts, sources, warnings; never charged).
`position`	`integer`	Position Rank of the paper in its query's results (1 = first).
`status`	`string`	Status ok \| not_found \| error \| skipped_invalid \| skipped_budget \| skipped_free_plan \| skipped_unmetered e.g. `"ok"`
`charged`	`boolean`	Charged Whether this record incurred a pay-per-event charge.
`chargedEvent`	`string`	Charged event Pay-per-event event charged for this record (present only when charged).
`chargedCount`	`integer`	Events charged Events actually charged when the charge came back short of the requested count (present only then; a warning explains it).
`source`	`string`	Source Paper rows: the sources searched (crossref+arxiv+pubmed+europepmc); lookups: the source that answered (crossref, arxiv, pubmed, europepmc, doi).
`fetchedAt`	`string`	Fetched at ISO timestamp.
`title`	`string \| null`	Title Paper title.
`authors`	`array`	Authors Author names as published, in order.
`year`	`integer \| null`	Year Publication year.
`publishedDate`	`string \| null`	Published Publication date: YYYY-MM-DD, YYYY-MM or YYYY (earliest of online and print).
`venue`	`string \| null`	Venue Journal, proceedings, book or preprint server (arXiv: the journal reference when published, else arXiv).
`publisher`	`string \| null`	Publisher Publisher (from Crossref or OpenAlex).
`type`	`string \| null`	Type journal-article, preprint, review, conference-paper, book, book-chapter, dataset, dissertation, report or other.
`doi`	`string \| null`	DOI DOI in lower case, without https://doi.org/.
`arxivId`	`string \| null`	arXiv id arXiv identifier without version (e.g. 1706.03762).
`pmid`	`string \| null`	PMID PubMed id.
`pmcid`	`string \| null`	PMCID PubMed Central id (free full text).
`openAlexId`	`string \| null`	OpenAlex id OpenAlex work id (W…), when OpenAlex was used.
`abstract`	`string \| null`	Abstract Abstract as plain text, when a source provides it.
`citationCount`	`integer \| null`	Citations Times cited: Crossref is-referenced-by-count, else OpenAlex or Europe PMC.
`openAccess`	`boolean \| null`	Open access Free to read: open-access licence, arXiv, PubMed Central or Europe PMC full text. null = unknown.
`pdfUrl`	`string \| null`	PDF Best free PDF link (arXiv, Europe PMC, publisher under an open licence).
`url`	`string \| null`	URL Landing page: where the paper can be read when it is open access, else the publisher's page.
`subjects`	`array`	Subjects MeSH headings and keywords (PubMed, Europe PMC), arXiv categories, OpenAlex topics.
`language`	`string \| null`	Language Language code, when a source states it.
`sources`	`array`	Found in Sources that returned this paper (crossref, pubmed, europepmc, arxiv, openalex, doi).
`idType`	`string`	Identifier type Lookups: doi, arxiv, pmid or pmcid.
`lookupId`	`string`	Identifier Lookups: the identifier in canonical form.
`query`	`string`	Query Summary records: the search query.
`resultsDelivered`	`integer`	Papers delivered Summary records: papers delivered (and charged) for the query.
`stopReason`	`string`	Stop reason Summary records: limit, free_plan, end, budget, aborted, deadline or error.
`sourcesSearched`	`array`	Sources searched Summary records: the sources asked.
`sourceTotals`	`object`	Hits per source Summary records: total matches each source reported (null when unknown or failed).
`duplicatesSkipped`	`integer`	Duplicates skipped Summary records: papers skipped because an earlier query of the run delivered them (not charged).
`error`	`object`	Error { code, message, provider?, status? } when status is not ok. Codes set by the runtime: budget_exhausted, free_plan_limit, unmetered, aborted, charge_failed, record_too_large, invalid_record, dataset_push_failed; upstream codes: http_error (with the source's HTTP status), timeout, network, tls_error, invalid_json, too_large, rate_limit, server, client, invalid_response, not_found; invalid_input for an identifier that is not a DOI, arXiv id, PMID or PMCID.
`warnings`	`array`	Warnings Notes about this record: sources that failed or were skipped, free-plan or budget limits, skipped duplicates.
`raw`	`object`	Raw Each source's unmodified record, keyed by source (only with includeRaw).

Example record

{
  "input": "crispr gene editing",
  "recordType": "paper",
  "position": 2,
  "title": "CRISPR-Cas9 system: A new-fangled dawn in gene editing",
  "authors": [
    "Darshana Gupta",
    "Oindrila Bhattacharjee",
    "Drishti Mandal"
  ],
  "year": 2019,
  "publishedDate": "2019-09",
  "venue": "Life Sciences",
  "publisher": "Elsevier BV",
  "type": "review",
  "doi": "10.1016/j.lfs.2019.116636",
  "arxivId": null,
  "pmid": "31295471",
  "pmcid": null,
  "openAlexId": null,
  "abstract": "Till date, only three techniques namely Zinc Finger Nuclease (ZFN), Transcription-Activator Like Effector Nucleases (TALEN) …",
  "citationCount": 299,
  "openAccess": null,
  "pdfUrl": null,
  "url": "https://linkinghub.elsevier.com/retrieve/pii/S0024320519305624",
  "subjects": [
    "Animals",
    "CRISPR-Cas Systems",
    "Gene Editing",
    "Genome editing"
  ],
  "language": "en",
  "sources": [
    "crossref",
    "pubmed"
  ],
  "status": "ok",
  "charged": true,
  "chargedEvent": "paper",
  "source": "crossref+pubmed+europepmc+arxiv",
  "fetchedAt": "2026-10-14T10:00:00.000Z"
}

FAQ

Questions

Which sources are searched?: Crossref (all disciplines, 170M+ DOI records), arXiv (preprints in physics, maths and computer science), PubMed (biomedical literature) and Europe PMC (life sciences, preprints and open-access full text). Results are merged: a paper found by several sources is one row, with the best fields from each (abstract from PubMed or arXiv, citation count from Crossref, open-access PDF from arXiv or Europe PMC). The `sources` field says where each paper was found.
Can I look up papers I already know?: Yes: put DOIs, arXiv ids, PMIDs or PMCIDs in `ids` (or links to them). Each returns one record and costs one `paper` event; identifiers that are not found are free. DOIs outside Crossref (DataCite, e.g. Zenodo or arXiv DOIs) are resolved through doi.org.
Do I get the full text?: No, the Actor returns metadata and abstracts. For open-access papers it gives the free PDF or full-text link (`pdfUrl`, `url`); pass those to a PDF or URL-to-markdown Actor for the text.

Machine-readable

This page for agents

/actors/academic-paper-search.json — this API as JSON
/actors/academic-paper-search.md — this API as Markdown
/catalog.json — the whole catalogue