Research Papers Search for RAG, from OpenAlex data

Independent tool. Not affiliated with, endorsed by or sponsored by OpenAlex or OurResearch; it uses their open (CC0) data through the public API.

Search 250M+ scholarly works in OpenAlex, the open catalog of research papers, or look up a list of DOIs, and get clean, citation-ready records built for RAG pipelines and research agents:

• title, authors with institutions (name, country, ROR) and ORCID, venue, year, DOI, citation count, FWCI, topics and keywords
• the abstract as plain text, rebuilt from OpenAlex's abstract_inverted_index (the API does not return abstracts as text)
• the best open-access PDF link and license
• ready-made citations in APA 7 and BibTeX
• a self-contained markdown card per paper (title, authors, venue, DOI, citations, abstract, APA), ready to embed in a vector database or paste into an LLM prompt
• deduplication by DOI (OpenAlex sometimes holds two records for one paper) and cursor paging, so large searches (up to 100,000 papers) are read 200 at a time without the 10,000-result cap of basic paging

Input

Field	Description
query	Words to find in titles, abstracts and full text. Supports AND, OR, NOT and "exact phrases".
dois	Optional list of DOIs (10.xxxx/... or https://doi.org/...). Lookup mode: the query and filters are ignored, duplicates are removed, unknown DOIs come back with status: "not_found" and are free.
yearFrom / yearTo	Publication year range.
openAccessOnly	Only papers free to read somewhere (default true).
types	Work types: article, preprint, review, book, book-chapter, dissertation, dataset, report...
fields	Research areas: OpenAlex's 26 fields (Computer Science, Medicine, Social Sciences...).
topics	OpenAlex topic ids (T10028) or names; names are matched to the closest topic and shown in the log.
minCitations	Only papers cited at least this many times.
language	ISO 639-1 code (en, pt, es...).
hasAbstract	Skip papers without an abstract (useful for RAG).
includeRetracted	Retracted papers are excluded unless this is on.
sort	relevance (needs a query), citations, newest or oldest.
maxResults	Stop after this many papers (default 25, up to 100,000).
includeMarkdown	Add the markdown card (default true).
contactEmail	Optional e-mail sent to OpenAlex as mailto (their "polite pool"). Not stored in results.
openalexApiKey	Optional OpenAlex API key for higher daily limits.

The default input searches open-access papers on "retrieval augmented generation" since 2023 and returns 25 papers in a few seconds.

Output

One dataset item per paper (views: Overview, Citations, Markdown for RAG):

{
  "status": "ok",
  "openalexId": "W2741809807",
  "doi": "10.7717/peerj.4375",
  "doiUrl": "https://doi.org/10.7717/peerj.4375",
  "title": "The state of OA: a large-scale analysis of the prevalence and impact of Open Access articles",
  "publicationYear": 2018,
  "type": "article",
  "authors": [
    {"name": "Juan Pablo Alperin", "orcid": "https://orcid.org/0000-0002-9344-7439", "position": "middle",
     "institutions": [{"name": "Simon Fraser University", "countryCode": "CA", "ror": "https://ror.org/0213rcc28", "type": "education"}]}
  ],
  "venue": "PeerJ",
  "volume": "6",
  "firstPage": "e4375",
  "citedByCount": 1000,
  "isOpenAccess": true,
  "oaStatus": "gold",
  "pdfUrl": "https://peerj.com/articles/4375.pdf",
  "license": "cc-by",
  "topics": [{"name": "scientometrics and bibliometrics research", "field": "Decision Sciences", "domain": "Social Sciences"}],
  "abstract": "Despite growing interest in Open Access (OA) to scholarly literature, there is an unmet need for ...",
  "citationApa": "Piwowar, H., Priem, J., Larivière, V., Alperin, J. P., & Haustein, S. (2018). The state of OA: ... PeerJ, 6, e4375. https://doi.org/10.7717/peerj.4375",
  "citationBibtex": "@article{piwowar2018state,\n  title = {The state of OA: ...},\n  author = {Piwowar, Heather and ...},\n  journal = {PeerJ},\n  year = {2018}, ...\n}",
  "markdown": "# The state of OA: ...\n\n- **Authors:** Heather Piwowar, Jason Priem, ...\n- **DOI:** https://doi.org/10.7717/peerj.4375\n...\n\n## Abstract\n\nDespite growing interest ...\n\n## Citation (APA)\n\n..."
}

(Example values are shortened.) Invalid DOIs return status: "error" and unknown DOIs status: "not_found"; neither is charged. The run's OUTPUT record has the totals (papers, duplicates skipped, OpenAlex requests, seconds).

Citations are generated from OpenAlex metadata on a best-effort basis: family names are split from display names (particles like van der and da are kept), and page ranges, volume and issue appear only when OpenAlex has them. Check them before formal submission.

How much does it cost?

Pay per event: US$ 0.001 per paper returned (paper event), i.e. US$ 1 per 1,000 papers. Invalid or unknown DOIs, duplicates and failed runs are free.

How it was set: the Actor runs with 256 MB and OpenAlex returns 200 papers per request, so 1,000 papers cost about US$ 0.0003 in Apify compute. Other OpenAlex Actors in the Store charge from about US$ 0.0004 to US$ 0.005 per paper, several with start fees; this one sits in the middle with no start fee while adding APA/BibTeX citations and a Markdown card. The run needs only its own default dataset and key-value store, so it works with Apify's limited permissions.

$0.001 per paper (paper event, pay per event). Charged once per paper returned. Invalid or unknown DOIs, duplicates and errors are free.

You get	Free tier	Bronze tier (10% off)	Gold tier (30% off)
100 papers	$0.10	$0.09	$0.07
1,000 papers	$1.00	$0.90	$0.70
10,000 papers	$10.00	$9.00	$7.00

Discounts follow your Apify subscription tier: Bronze 10%, Silver 20%, Gold and above 30%. No subscription to this Actor and no charge for platform usage: you pay only for results. The Apify Free plan includes $5 of monthly credit, enough for about 5,000 papers at the list price. You can cap the cost of any run with Maximum cost per run; the Actor stops when it is reached.

Data source

OpenAlex, a free and open index of scholarly works run by the nonprofit OurResearch. OpenAlex data is released under CC0 (public domain), so it can be reused for any purpose, commercial included; attribution is appreciated and given here. The Actor calls the official REST API (api.openalex.org) and follows its rules: mailto for the polite pool, cursor paging, select to request only needed fields, and back-off on 429. The free tier has a daily credit budget (each list request of up to 200 papers costs 10 credits); if it runs out, the run stops cleanly with a message, and you can pass your own openalexApiKey. If OpenAlex cannot be reached at all, the run fails with a clear message after five retries instead of returning partial data silently.

Abstracts are reconstructed from the inverted index OpenAlex publishes; coverage is higher for recent works (about 60% of 2022 works have one). Authors' names, ORCID and institutions are public authorship metadata; raw affiliation strings (which sometimes contain e-mail addresses) are not included.

More questions

Does it return the full text of the papers?

No. It returns metadata, the abstract and the best open-access pdfUrl with its license. To put full papers in a RAG index, download the PDFs from pdfUrl (filter with openAccessOnly, on by default) and convert them, for example with our Documents to Markdown Actor. The markdown card is meant as a compact, citable record per paper.

Why do some papers have no abstract?

OpenAlex publishes abstracts only as an inverted index, and not every work has one; coverage is higher for recent works. The Actor rebuilds the text when the index exists, otherwise abstract is null and the Markdown card says no abstract is available. Turn on hasAbstract to keep only papers with an abstract, which is usually what a RAG index needs.

How do I chunk the results for a vector database?

Each paper is already one self-contained chunk: the markdown card holds the title, up to 10 authors with institutions, venue and year, DOI, citations, open-access link, topics, abstract and APA citation. Embed markdown (or title + abstract) and keep doi, publicationYear, citedByCount and topics as metadata for filtering and citing.

How fresh is the data?

Every run queries the live OpenAlex API, so results are as current as OpenAlex, and citation counts change over time. Use sort: "newest" with a yearFrom to monitor a topic on a schedule.

When only the year is known, what is the publication date?

OpenAlex then sets publicationDate to January 1 of that year. Use publicationYear when you need a reliable value.

How do DOI lookups work?

Put DOIs in dois (bare 10.xxxx/... or https://doi.org/...). The query and filters are ignored in this mode, repeated DOIs are looked up once, inputs that are not DOIs come back with status: "error", and DOIs that OpenAlex does not have come back with status: "not_found". Both are free.

What are the limits?

Up to 100,000 papers per run, read 200 per request with cursor paging. OpenAlex's free tier has a daily credit budget; when it runs out the run stops cleanly with a message, and passing your own openalexApiKey raises the limit. Without a query, results are sorted by most cited; relevance needs a query.

Can I combine keywords with fields and topics?

Yes. query supports AND, OR, NOT and "exact phrases", and is combined with every filter: years, open access, types, fields (OpenAlex's 26 research areas), topics (ids like T10028 or names, matched to the closest topic and shown in the log), language, minimum citations and retracted papers (excluded by default).

Use it via API

Run it from any language and get the results as JSON. Replace <YOUR_API_TOKEN> with your Apify API token.

curl -X POST "https://api.apify.com/v2/acts/br_data_hub~openalex-research-search/run-sync-get-dataset-items?token=<YOUR_API_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{"query": "retrieval augmented generation"}'

from apify_client import ApifyClient

client = ApifyClient("<YOUR_API_TOKEN>")
run = client.actor("br_data_hub/openalex-research-search").call(run_input={'query': 'retrieval augmented generation'})
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item)

You can also schedule runs, send results to webhooks, or connect the Actor to Make, Zapier, n8n, Google Sheets and other integrations.

Use it with AI agents (MCP)

This Actor is ready for AI agents: typed input and output schemas, pay-per-event pricing, limited permissions and short runs.

• Apify MCP server: add https://mcp.apify.com/?tools=br_data_hub/openalex-research-search to Claude, ChatGPT, Cursor, VS Code or any MCP client, and the agent can call this Actor as a tool.
• Agent frameworks: call it from LangChain, LlamaIndex, CrewAI or the OpenAI Agents SDK through the Apify integrations, or with the API call above.
• Results come as structured JSON with stable field names, documented in the Output tab.

FAQ

Where does the data come from?

OpenAlex, a free and open index of scholarly works run by the nonprofit OurResearch. OpenAlex data is released under CC0 (public domain), so it can be reused for any purpose, commercial included; attribution is appreciated and given here. The Actor calls the official REST API (api.openalex.org) and follows its rules: mailto for the polite pool, cursor paging, select to request only needed fields, and back-off on 429. The free tier has a daily credit budget (each list request of up to 200 papers costs 10 credits); if it runs out, the run stops cleanly with a message, and you can pass your own openalexApiKey. If OpenAlex cannot be reached at all, the run fails with a clear message after five retries instead of returning partial data silently.

How much does it cost?

$0.001 per paper. Discounts follow your Apify subscription tier: Bronze 10%, Silver 20%, Gold and above 30%. See How much does it cost? above for examples.

Do I pay for errors or empty results?

No. Charged once per paper returned. Invalid or unknown DOIs, duplicates and errors are free.

Can I run it on a schedule or connect it to other tools?

Yes. Use Apify schedules to run it daily or weekly, webhooks to get notified, and integrations such as Make, Zapier, n8n and Google Sheets. Everything also works through the API.

Can AI agents use it?

Yes. It works as a tool in the Apify MCP server and in agent frameworks, and AI agents can pay per event. See Use it with AI agents (MCP).

Is it legal to use this data?

The Actor reads only public sources. You are responsible for how you use the data and for complying with the source's terms and with privacy laws such as GDPR.

I found a bug or need a feature. What should I do?

Open an issue in the Issues tab. We answer within 14 days, usually much sooner.

More Actors from BR Data Hub

Actor	What it does
Documents to Markdown for RAG (Docling)	Convert PDF and DOCX files to clean Markdown with Docling: page count, detected tables, metadata and optional
Website to llms.txt Generator	Crawl one website (robots.txt and sitemap aware), generate spec-compliant llms.txt and llms-full.txt files, au
RSS & Atom Feed Reader with Full Text	Read RSS, Atom and JSON Feed (or find the feed of any website) and get clean items with ISO dates, full articl
Brazil CNPJ Lookup API (Consulta CNPJ)	Look up Brazilian companies by CNPJ: registration status, partners (masked CPF), activities, tax regime and sc

See all our Actors at apify.com/br_data_hub.

Usar em português

Busque artigos científicos a partir dos dados do OpenAlex (mais de 250 milhões de trabalhos, dados em domínio público, CC0) por palavra-chave, ano, acesso aberto, tipo, área, tópico, idioma e número mínimo de citações, ou informe uma lista de DOIs. Cada artigo vem com autores e instituições, periódico, ano, DOI, citações, tópicos, link do PDF aberto, resumo em texto (reconstruído do índice invertido do OpenAlex), citação pronta em APA e BibTeX e um campo markdown pronto para RAG e agentes de pesquisa. Resultados repetidos pelo mesmo DOI são removidos. A cobrança é por artigo entregue (paper, US$ 0,001); DOIs inválidos ou não encontrados não são cobrados. Ferramenta independente, sem afiliação com o OpenAlex ou a OurResearch.

Quanto custa: $0.001 por resultado (paper); 1.000 resultados saem por $1.00. Desconto pelo nível da assinatura Apify: Bronze 10%, Silver 20%, Gold ou acima 30%. Erros e consultas sem resultado não são cobrados. Use também pela API ou por agentes de IA no servidor MCP da Apify (https://mcp.apify.com/?tools=br_data_hub/openalex-research-search).