PubMed Scraper — Abstracts + MeSH, 11 Fields, No API Key

The Mine Works: 544 users across 100 public Actors, >99% run success. Pay only for results delivered. Browse all Actors.

Search PubMed and get 11 fields per article: PMID, title, full abstract, author list, journal, ISSN, publication year, DOI, MeSH terms, the PubMed link and a capture timestamp. Filter by any PubMed query syntax plus a date range, and pull up to 10,000 articles per run.

Pure HTTP against NCBI's official E-utilities. No API key required, no login, no browser — and if you do have an NCBI key, supply it to run roughly three times faster.

Why use this PubMed scraper

• MeSH terms parsed out. mesh_terms is the controlled-vocabulary index that makes biomedical literature searchable by concept rather than keyword. It is the field most PubMed exports lose.
• Full abstracts, assembled correctly. Structured abstracts arrive from NCBI as multiple labelled sections; this Actor joins them into one clean text field.
• Native PubMed query syntax. query is passed straight through, so field tags and boolean operators work exactly as they do on pubmed.ncbi.nlm.nih.gov.
• Optional NCBI key for speed. Without a key the Actor paces at roughly three requests per second; with one it paces at nine, which is NCBI's documented allowance.
• An upstream outage is reported as an outage. When NCBI's own backend fails it returns HTTP 200 with an error inside the body. This Actor surfaces that as an upstream problem instead of reporting "no results" — those are not the same thing and only one is worth retrying.

Search a topic with a date floor

{
  "query": "GLP-1 receptor agonist diabetes",
  "dateFrom": "2020/01/01",
  "maxResults": 500
}

Dates use PubMed's own YYYY/MM/DD format and are compiled into a [dp] date-of-publication filter.

Use PubMed field tags for a precise search

{
  "query": "semaglutide[Title/Abstract] AND cardiovascular[MeSH Terms]",
  "maxResults": 300
}

Anything that works in the PubMed search box works here: [Title/Abstract], [MeSH Terms], [Author], [Journal] and boolean operators.

Build a literature corpus for RAG

{
  "query": "alzheimer disease biomarkers",
  "dateFrom": "2018/01/01",
  "maxResults": 5000
}

Title, full abstract, MeSH terms and DOI on every row is a complete chunk-and-embed input set, with a stable PMID as the citation key.

Run a large pull with an NCBI key

{
  "query": "CRISPR therapeutic",
  "dateFrom": "2015/01/01",
  "maxResults": 10000,
  "ncbiApiKey": "YOUR_NCBI_KEY"
}

An NCBI key is free from an NCBI account and raises the request rate from three to nine per second, which materially shortens a 10,000-article run.

What data does the PubMed scraper return

{
  "pmid": "37890123",
  "title": "Cardiovascular outcomes with GLP-1 receptor agonists: a meta-analysis",
  "abstract": "BACKGROUND: Glucagon-like peptide-1 receptor agonists... METHODS: We searched...",
  "authors": ["Jane A Doe", "John B Smith"],
  "journal": "The Lancet Diabetes & Endocrinology",
  "issn": "2213-8587",
  "year": "2023",
  "doi": "10.1016/S2213-8587(23)00123-4",
  "mesh_terms": ["Diabetes Mellitus, Type 2", "Glucagon-Like Peptide-1 Receptor", "Cardiovascular Diseases"],
  "url": "https://pubmed.ncbi.nlm.nih.gov/37890123/",
  "scraped_at": "2026-08-01T20:00:00.000Z"
}

Field	Description
pmid	PubMed identifier
title, abstract	Article title and full abstract text
authors[]	Author names, forename plus surname
journal, issn, year	Publication venue and year
doi	DOI where the record carries one
mesh_terms[]	MeSH descriptor terms assigned by NLM indexers
url, scraped_at	PubMed link and capture timestamp

Very recent articles are often not yet MeSH-indexed, so mesh_terms can be empty on new records. Some records carry no abstract or DOI at all; those fields come back empty rather than filled in.

How the scraper works without an API key

NCBI's E-utilities are public. This Actor runs esearch to collect PMIDs for your query, then efetch in batches of 50 to pull the full XML records, and parses title, abstract sections, authors, journal, DOI and MeSH headings out of each. Requests are paced at roughly three per second without a key and nine per second with one, matching NCBI's published limits.

What can you build with PubMed data

Systematic reviews and meta-analyses. A reproducible, dated corpus with abstracts and MeSH terms ready to screen.

Biomedical RAG and AI training corpora. Clean abstracts with a stable citation key and concept-level MeSH tags.

Competitive and scientific monitoring. A scheduled query on a molecule, mechanism or competitor's research output.

Bibliometrics. Publication volume by journal, year or topic, joinable to Crossref citation counts via DOI.

How much does it cost to scrape PubMed

Pay per article delivered: $0.0025 on the Apify Free plan, $0.0015 on Gold and above. Queries that match nothing, and runs that stop because NCBI's backend is down, are never charged.

How do I use PubMed in Claude or ChatGPT

https://mcp.apify.com/?tools=themineworks/pubmed-ncbi-scraper

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: 'YOUR_APIFY_TOKEN' });

const run = await client.actor('themineworks/pubmed-ncbi-scraper').call({
  query: 'GLP-1 receptor agonist diabetes',
  dateFrom: '2020/01/01',
  maxResults: 500,
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();

PubMed scraper FAQ

Do I need an NCBI API key? No. It is optional and only raises the request rate from three to nine per second.

What query syntax is supported? All of PubMed's. Field tags and boolean operators are passed straight through to esearch.

Why does an article have no MeSH terms? NLM indexing lags publication, so recent articles are often not yet indexed.

What date format do I use? PubMed's own YYYY/MM/DD.

Can I get full text? No. PubMed publishes abstracts; the DOI on each row is your route to the publisher's full text.

Complete your biomedical research pipeline

• Crossref Scraper — citation counts for the same DOIs.
• OpenCitations Scraper — the citation graph around any DOI.
• NIH RePORTER Grants Scraper — the funding behind the research.

Found a bug or want a field added? Open an issue on the Actor's Apify Console page.

Related guides

• PubMed Scraper: Search 36M Biomedical Articles, Abstracts, and MeSH Terms via API

Last verified: 2026-08