Academic Papers Scraper: OpenAlex Research Paper Alerts

Academic papers scraper for OpenAlex research papers: returns paper metadata as clean structured data, one record per paper with title, DOI, publication date, venue, work type, open access status and the free-to-read link, authors with their institutions and ORCIDs, institutions with ROR IDs, research topics and concepts, citation count, the abstract where its licence allows, source links and a licence attribution line.

Search by keyword, author, institution or topic, filter by publication date and open access, sort by newest or most cited, and switch on onlyNew to get only the papers that appeared since your last run. Schedule it and you have a literature alert feed.

The data comes from OpenAlex, the open catalogue of over 300 million scholarly works, released as CC0. You can add arXiv details for preprints and Crossref publisher, licence and funder details.

Use it to:

• Get a daily or weekly research paper alert for your topics, authors or institution.
• Build a reading list for a literature review, sorted by newest or most cited.
• Track what a university, lab or company publishes.
• Find the free to read copy of a paper through its open access link.
• Build AI or analytics datasets with CC0 metadata and abstracts.

What you get

Field	Example
title, doi	Experimental reasoning in large language models, 10.64898/2026.10.20.760214
publicationDate, publicationYear	2026-10-21, 2026
venue, venueType, publisher, issn	bioRxiv (Cold Spring Harbor Laboratory), repository, Cold Spring Harbor Laboratory
type, language	article, preprint, review, book-chapter, dataset and more
isOpenAccess, oaStatus, oaUrl, workLicence	true, green, link to the free copy, cc-by
authorNames, authorCount, authors	Names in byline order; per author: position, ORCID, OpenAlex ID, institutions, countries
institutionNames, institutions, countries	Per institution: name, OpenAlex ID, ROR, country, type
primaryTopic, topics, concepts, keywords	Topic with subfield, field and domain
citationCount, referencesCount, fwci, isRetracted	Citations so far, references, field-weighted citation impact
abstract, abstractSource	Abstract text from OpenAlex or arXiv (both CC0)
openAlexId, pmid, arxivId, landingPageUrl, sourceLinks	Links to OpenAlex, DOI, publisher page, open access copy, arXiv and PubMed
arxiv	With arXiv enrichment: categories, version, dates, comment, journal reference
crossref	With Crossref enrichment: publisher, licences, funders and award numbers, ISSNs, subjects
matchedQueries, isNew, firstSeenAt	Which of your searches found it, and whether it is new since the previous run
licence, dataSources, primarySource	Licence and attribution for the metadata in this record

The Output tab has four views: Papers, Open access, Citations and topics and Abstracts.

Who it is for

• Researchers and librarians who want a weekly list of new papers on their topics, by their collaborators or from their institution.
• R&D, competitive intelligence and patent teams tracking what universities and companies publish.
• Research offices and funders monitoring output, open access share and citations.
• AI and data teams building literature datasets with clear CC0 licensing.
• Science journalists and analysts following a field as it moves.

Quick start

1. Click Start with the example input. You get the newest papers about large language models.
2. Replace the search with your own words, or add authors (ORCID works best), institutions (ROR ID works best) or topics.
3. Narrow it down with dates, Published in the last N days or Open access only.
4. For alerts, switch on Only papers not seen in previous runs, schedule the Actor daily or weekly, and send the dataset to email, Slack, Google Sheets or your database with an Apify integration or webhook.

Input

Field	What it does	Default
searchQueries	One search per entry. Words are combined with AND; use quotes for phrases and OR or NOT in capitals	example query
searchIn	titleAbstractKeywords, titleAndAbstract, title or everything (adds full text where OpenAlex has it)	titleAbstractKeywords
authors	ORCIDs, OpenAlex author IDs or names	none
institutions	ROR IDs, OpenAlex institution IDs or names. Includes departments and labs	none
topics	OpenAlex topic IDs or topic names	none
fromDate, toDate	Publication date window, YYYY-MM-DD	none
publishedWithinDays	Only papers published in the last N days	0 (off)
openAccessOnly	Only papers that are free to read	false
sort	newest, mostCited or relevance	newest
maxItems	Maximum paper records in total, shared between searches	100
onlyNew	Only papers not returned by a previous run with the same search and filters	false
includeAbstracts	Add CC0 abstracts from OpenAlex or arXiv	true
enrichWithArxiv	Add arXiv details for papers on arXiv	false
enrichWithCrossref	Add Crossref publisher, licence and funder details	false
openAlexApiKey	Optional free OpenAlex API key (stored as a secret)	none

Example input for a weekly alert:

{
  "searchQueries": ["\"solid-state battery\" electrolyte", "sodium-ion battery"],
  "institutions": ["https://ror.org/042nb2s44"],
  "publishedWithinDays": 14,
  "openAccessOnly": false,
  "sort": "newest",
  "onlyNew": true,
  "maxItems": 200,
  "openAlexApiKey": "YOUR_FREE_KEY"
}

Names, IDs and how matching works

Names are looked up in OpenAlex and the best match is used; the match is listed in the RUN_SUMMARY record under resolvedFilters, so you can check it. Common names can match the wrong person, so use an ORCID for authors and a ROR ID for institutions when you can. Several authors, institutions or topics are combined with OR inside each group, and the groups are combined with AND.

Do I need an OpenAlex API key?

No. Without a key the Actor reads the first page of results for each search, up to 100 papers, which is plenty for the example, for small searches and for most daily alerts.

A free key from OpenAlex (create an account at openalex.org, then Settings, API key) gives you:

• Paging through all results, up to 5,000 per search per run.
• Your own daily allowance of roughly 1,000 searches, instead of a small allowance shared by everyone on the same IP address.

When the shared allowance is used up and no key is given, the Actor still returns results: it searches Crossref for matching DOIs and then reads each paper from OpenAlex as a single record, which OpenAlex does not charge for. These records show discoveredVia: "crossref-search". Matching in this mode is looser (author and institution names are matched as text), and with "Newest first" it looks back 90 days unless you set a start date. The run summary says when this happened. With a key, the Actor uses OpenAlex search directly.

Output example

{
  "title": "Experimental reasoning in large language models",
  "doi": "10.64898/2026.10.20.760214",
  "publicationDate": "2026-10-21",
  "publicationYear": 2026,
  "venue": "bioRxiv (Cold Spring Harbor Laboratory)",
  "venueType": "repository",
  "publisher": "Cold Spring Harbor Laboratory",
  "type": "preprint",
  "isOpenAccess": true,
  "oaStatus": "green",
  "oaUrl": "https://www.biorxiv.org/content/biorxiv/early/2026/10/21/2026.10.20.760214.full.pdf",
  "workLicence": "cc-by",
  "authorNames": "Jeongyheon Kim; Philip Romero",
  "authorCount": 2,
  "authors": [
    {
      "name": "Jeongyheon Kim",
      "position": "first",
      "orcid": null,
      "openAlexId": "A5126074038",
      "institutions": ["Duke University"],
      "countries": ["US"],
      "isCorresponding": false
    }
  ],
  "institutionNames": "Duke University",
  "institutions": [
    {"name": "Duke University", "openAlexId": "I170897317", "ror": "https://ror.org/00py81415", "countryCode": "US", "type": "education"}
  ],
  "primaryTopic": "AI-based Problem Solving and Planning",
  "keywords": ["experimental reasoning", "large language models", "laboratory automation", "AI agents"],
  "citationCount": 0,
  "abstract": "Scientific reasoning requires interaction with the physical world: designing experiments, interpreting outcomes, and deciding what experiment should come next. ...",
  "abstractSource": "openalex",
  "openAlexId": "W7222941417",
  "sourceLinks": {
    "openAlex": "https://openalex.org/W7222941417",
    "doi": "https://doi.org/10.64898/2026.10.20.760214",
    "openAccess": "https://www.biorxiv.org/content/biorxiv/early/2026/10/21/2026.10.20.760214.full.pdf"
  },
  "matchedQueries": ["large language models"],
  "isNew": true,
  "primarySource": "openalex",
  "dataSources": ["openalex"],
  "licence": "Metadata from OpenAlex (openalex.org), CC0 1.0 public domain dedication. Abstract from OpenAlex, CC0 1.0. Linked full texts keep their own licences."
}

A run report is saved in the key-value store as RUN_SUMMARY: records returned, how each author, institution or topic was matched, the date window used, pages read, enrichment counts and any notes.

Pricing

This Actor uses pay per event pricing, with platform usage included:

• $1.50 per 1,000 paper records (paper-record event, $0.0015 each)

You only pay for paper records saved to your dataset, so filters, onlyNew and maxItems keep costs down. Enrichment from arXiv and Crossref is included in the price. Set a maximum cost per run in the run options and the Actor stops cleanly when it is reached. If you use your own OpenAlex API key, any OpenAlex usage beyond its free daily allowance is between you and OpenAlex.

How "only new papers" works

When onlyNew is on, the Actor remembers which papers it has already returned for this exact search and filters, in a key-value store in your account called agents001-academic-research-feed-state. The next run only returns papers that are not on that list. The first run returns everything that matches. Sorting, maxItems and enrichment settings do not reset the history; changing the search or filters starts a fresh one. The store holds paper IDs and dates only.

With onlyNew off, the Actor does not open or update this store, so every record has isNew set to true. If a run stops with a message that it cannot open the store (for example because another Actor in your account already uses that name), enter a different name in State store name and run again.

For alerts, combine onlyNew with publishedWithinDays (for example 14) and "Newest first". Papers are sometimes added to OpenAlex weeks after their publication date, and the window catches those late arrivals.

Data sources, licences and privacy

• OpenAlex (openalex.org): all metadata and abstracts are released under CC0 1.0. Abstracts are rebuilt from OpenAlex's word index. Cite OpenAlex: Priem, J., Piwowar, H., and Orr, R. (2022). OpenAlex: A fully-open index of scholarly works, authors, venues, institutions, and concepts. arXiv:2205.01833.
• arXiv (arxiv.org, optional): descriptive metadata including abstracts is CC0 1.0. The Actor reads metadata only, keeps to one request every three seconds, and links to arXiv abstract pages. It never downloads, stores or serves arXiv PDFs.
• Crossref (crossref.org, optional and fallback): bibliographic metadata, free to reuse. Crossref abstracts are never included, because publishers or authors may hold copyright in them.
• Every record carries a licence field naming the sources used for it. Full texts reached through oaUrl or other links keep their own licences (see workLicence).
• Authors: names, ORCIDs and institutional affiliations are public scholarly authorship data and are included. Email addresses and raw affiliation text (which can contain emails) are never output.

FAQ

Why did I get fewer papers than maxItems? Without an API key only the first page (up to 100 papers) of each search is read, and with several searches maxItems is shared between them. With onlyNew, papers returned before are skipped. Add a free OpenAlex key to page further.

Why is the abstract empty for some papers? OpenAlex has no abstract for some works, often books, chapters and some publishers' articles. Switch on arXiv enrichment to fill it for preprints that are on arXiv.

Why do some papers have a publication date in the future? Journals often date issues ahead. With "Newest first" and no end date, the Actor stops at today. Set toDate if you want them.

How fresh is the data? OpenAlex adds tens of thousands of works a day, usually within days of publication. Citation counts are updated continuously.

How much does a typical run cost? $1.50 per 1,000 paper records, with arXiv and Crossref enrichment included. A run with the default limit of 100 papers costs at most $0.15.

Can I schedule it and get only new papers? Yes. Schedule it in Apify daily or weekly and switch on Only papers not seen in previous runs (onlyNew). Each run then returns only papers not returned before for the same search and filters.

Does it need a login or API key? No login is needed. An OpenAlex API key is optional and free; see Do I need an OpenAlex API key? above for what it adds.

What formats can I export? Apify gives you the dataset as JSON, CSV or Excel, and through the Apify API or an integration.

Is it legal? Where does the data come from? Metadata and abstracts come from OpenAlex under CC0 1.0, with optional arXiv metadata (CC0) and Crossref bibliographic metadata. Crossref abstracts and arXiv PDFs are never included. Every record carries a licence field naming its sources.

Something looks wrong. What should I do? Open an issue on the Actor's Issues tab with the run link and the input.

Disclaimer

This Actor is independent and is not affiliated with, endorsed by or sponsored by OpenAlex, OurResearch, arXiv, Cornell University, Crossref or any publisher. Source names are used only to describe where the data comes from. Check the original paper before relying on its details.