Research Paper Alerts: PubMed, Europe PMC and OpenAlex Watch

Get the new research papers and preprints for your keywords, one clean row per paper, from PubMed, Europe PMC, bioRxiv, medRxiv, arXiv and OpenAlex in a single run. Each watch query is searched in two open indexes: Europe PMC, which carries PubMed, PubMed Central and life-science preprints, and OpenAlex, which adds arXiv and every other field. A paper that both indexes hold comes back once, with its DOI, PubMed ID, abstract, journal or preprint server, a free full-text link when one exists, and the list of places it was found. Turn on monitoring and put the run on a schedule to get only the papers that are new since your last run, charged per paper and free on a quiet day.

Who uses it: researchers and lab managers who want a keyword alert without five separate email digests, biotech business-development and competitive-intelligence teams tracking a target, a drug or a method, medical-affairs and pharmacovigilance teams watching a compound, and builders who feed new abstracts to a summariser or a Slack channel.

Related actors: for a one-off literature search with citation counts and author affiliations use OpenAlex Work Search; to resolve a list of DOIs use Crossref DOI Lookup. This actor is the alert: newest first, de-duplicated, with a memory.

How to use

1. In the Apify Console. Open the actor page and click Start — the queries field is already pre-filled with a working example. Results land in the run's dataset as soon as each item is found.
2. Via the API. Call it directly with a POST request — no Console needed once you have an API token:

   curl "https://api.apify.com/v2/acts/accountable_eel~research-literature-watch/run-sync-get-dataset-items?token=<YOUR_TOKEN>" \
     -X POST \
     -H "Content-Type: application/json" \
     -d '{"queries":["llm"]}'
   

3. On a schedule. Save this actor as an Apify Task with the input you want, then add a Schedule (hourly, daily, weekly) so it runs on its own — no server of your own required.

Input

{
  "queries": [
    "llm"
  ]
}

One watch per line, up to 20. Plain words are all required (CRISPR base editing); put a phrase in quotes ("large language model"); AND, OR, NOT and parentheses work. Each line is its own watch with its own memory. Accepted formats: llm, "large language model" AND radiology, CRISPR base editing, (semaglutide OR tirzepatide) AND obesity.

Each line of queries is its own watch. Words are matched in the title and abstract by default, which keeps an alert on topic; switch matchIn to the full text for a broad sweep. sinceDays (default 30) is the publication window and maxItems (default 25, up to 100) caps the papers per query, newest first. Include and exclude keywords are applied before anything is remembered or charged. Uploads to general-purpose repositories (Zenodo, figshare, Kaggle) are left out: on a keyword they are mostly datasets, slides and self-published notes, not papers.

OpenAlex gives anonymous callers a small daily allowance per address. When it is used up, the run still returns the Europe PMC papers and names OpenAlex in the sourcesFailed column. Add your own free OpenAlex API key in openAlexApiKey to always include it. No key is needed to run.

Monitoring. With onlyNewSinceLastRun on, the first run returns the newest papers and remembers everything it saw; every later run returns only papers it has not returned before. A paper is recognised by its DOI, its PubMed ID or its title, so a record that gains a DOI a week later is not reported twice. A run with nothing new returns one row with paperCount 0 and is not charged.

Output

One row per paper, for example:

query	found	status	searchQuery	paperCount	monitorStatus	truncated	sourcesChecked	sourcesFailed	papers	title	authors	firstAuthor	authorCount	journal	publishedAt	publicationType	isPreprint	preprintServer	doi	pmid	pmcid	openalexId	abstract	isOpenAccess	fullTextUrl	foundIn	paperUrl	sourceUrl	isNew	firstSeenAt	scrapedAt
llm	true	OK							<all papers (full list)>		<authors>																		<is this paper new?>		1970-01-01T00:00:00.000Z

Three real rows from a run on llm (2026-10-11), trimmed to the main columns:

title	journal	publishedAt	publicationType	doi	foundIn
A multi-model LLM framework for AI-assisted psychological scale development	BMC Psychology	2026-10-11	journal-article	10.1186/s40359-026-05759-w	OpenAlex
DBCLS BioHackathon 2026 report: An interactive, harmonised BioSample metadata resource	BioHackrXiv	2026-10-10	preprint	10.37044/osf.io/hm9fn_v1	Europe PMC, BioHackrXiv
Physics-guided thermodynamic screening of H2 adsorption on alkaline-earth-modified anatase TiO2 nanoparticles	RSC advances	2026-10-09	journal-article	(none yet)	Europe PMC, PubMed Central

A query that matches nothing comes back as one row with "found": false and is never charged. A value an index does not hold is null, never 0 or an empty string: a brand-new PubMed Central record often has no DOI or abstract yet. Author names are returned as published; affiliations, emails and ORCID iDs are not. A preprint and its later journal version have two DOIs and are two rows.

Pricing

Pay-per-event. A flat per-run fee covers session/proxy warmup; you're billed per item only when data is actually found and returned — see .actor/pay_per_event.json for exact prices. A miss is never charged.

Example runs

Paste any of these into the Input tab and press Start. All three return rows with no key, no token and no account of your own beyond an Apify login.

The newest papers on one keyword. The default shape.

{
  "queries": ["llm"],
  "maxItems": 20
}

Two watches in one run, life-science indexes only, last 14 days.

{
  "queries": ["CRISPR base editing", "(semaglutide OR tirzepatide) AND obesity"],
  "sources": "europePmc",
  "sinceDays": 14,
  "maxItems": 25
}

A daily alert: only papers new since the last run, study protocols left out.

{
  "queries": ["\"large language model\" AND radiology"],
  "excludeKeywords": ["study protocol"],
  "onlyNewSinceLastRun": true,
  "deltaName": "radiology-llm"
}

Use it from Clay, n8n, Make, or an AI agent

This actor runs synchronously over plain HTTP — call it directly from a script, a workflow tool, or an AI agent, no Apify Console needed once you have an API token.

curl "https://api.apify.com/v2/acts/accountable_eel~research-literature-watch/run-sync-get-dataset-items?token=<YOUR_TOKEN>" \
  -X POST \
  -H "Content-Type: application/json" \
  -d '{"queries":["llm"]}'

n8n. Add an HTTP Request node: Method POST, URL https://api.apify.com/v2/acts/accountable_eel~research-literature-watch/run-sync-get-dataset-items?token=<YOUR_TOKEN>, Body Content Type JSON, JSON Body {"queries":["llm"]} (swap in an expression from an earlier node for a real value).

Clay. Add an "HTTP API" column: Method POST, URL https://api.apify.com/v2/acts/accountable_eel~research-literature-watch/run-sync-get-dataset-items?token=<YOUR_TOKEN>, Body {"queries":["{{watch query}}"]}, mapping the row's watch query into the queries array.

MCP. In Claude, Cursor, or any MCP client with the Apify MCP server, ask for "new research papers alert by keyword" — the agent will find and run this actor.