Get Singleton

Use GET /{entity_type}/{id} to retrieve a single entity. The response is the full entity object.

curl "https://api.openalex.org/works/W2741809807"

Cost: Free, for every entity type: works, authors, sources, institutions, topics, publishers, funders and awards, and the vocabularies too (/keywords/climate-change, /countries/US, /work-types/article, /locations/doi:10.7717/peerj.4375).

A random record: /works/random (also /authors/random, /sources/random and the other main entities) returns one record chosen at random. It runs a query to choose it, so it costs what a list costs: 1 credit.

Fetching many? One list request returns up to 100 records by ID, much faster than 100 single requests. It costs 1 credit for the whole list (see OR filters):

curl "https://api.openalex.org/works?filter=ids.openalex:W2741809807|W2100837269&per_page=100"

OpenAlex IDs

Every entity has a unique OpenAlex ID with a letter prefix indicating the type:

PrefixEntity
WWork
AAuthor
SSource
IInstitution
TTopic
KKeyword
PPublisher
FFunder
GAward

You can use the full URL or just the key:

# Full URL
curl "https://api.openalex.org/works/https://openalex.org/W2741809807"

# Key only (recommended)
curl "https://api.openalex.org/works/W2741809807"

IDs are case-insensitive: W2741809807 and w2741809807 are equivalent.

External ID lookups

You can look up entities using external identifiers. Pass the full external ID URL or use shorthand notation.

# DOI
curl "https://api.openalex.org/works/https://doi.org/10.7717/peerj.4375"

# ORCID
curl "https://api.openalex.org/authors/https://orcid.org/0000-0001-6187-6610"

# ROR
curl "https://api.openalex.org/institutions/https://ror.org/02y3ad647"

# ISSN
curl "https://api.openalex.org/sources/issn:0028-0836"

# PMID
curl "https://api.openalex.org/works/pmid:29456894"

# arXiv
curl "https://api.openalex.org/works/arxiv:2305.18290"

Short-form IDs

For convenience, you can use prefix notation instead of full URLs:

curl "https://api.openalex.org/works/doi:10.7717/peerj.4375"
curl "https://api.openalex.org/works/pmid:29456894"

Supported external IDs

Each entity type has a primary (canonical) external identifier, and most accept a few more. Use the prefix form (ror:02y3ad647) or, for IDs that have one, the full URL (https://ror.org/02y3ad647, https://orcid.org/..., https://doi.org/...):

EntityCanonical IDLookup forms
WorksDOIdoi: (or a bare DOI, 10.7717/peerj.4375), arxiv: (or an arxiv.org URL), pmid: (or the PubMed URL https://pubmed.ncbi.nlm.nih.gov/29456894), pmcid:, mag:
AuthorsORCIDorcid:
SourcesISSN-Lissn_l:, issn:
InstitutionsRORror:, wikidata:
FundersROR or Funder DOIror:, doi: (the Funder DOI, e.g. doi:10.13039/100000001), crossref: (the Funder Registry id, e.g. crossref:100000001), wikidata:
PublishersWikidata IDwikidata:, ror:
Concepts (deprecated)Wikidata IDwikidata:

Prefixes are case-insensitive: DOI:10.7717/peerj.4375, PMID:29456894 and arXiv:2305.18290 work too. A DOI pasted with the punctuation of its sentence still finds its work: doi:10.7717/peerj.4375. (or with a trailing , or ;) answers like doi:10.7717/peerj.4375.

An arXiv id finds the work that holds arXiv’s DOI for that paper (10.48550/arXiv.2305.18290), so arxiv:2305.18290, https://arxiv.org/abs/2305.18290v2 and doi:10.48550/arxiv.2305.18290 return the same work. That work may be the published article, when the preprint has been merged with it. A version suffix (v2) is ignored, and old-style ids like arxiv:hep-th/9711200 work too.

Wikidata IDs work as wikidata:Q746413 or either Wikidata URL (https://www.wikidata.org/wiki/Q746413 or .../entity/Q746413). Topics have no external ID to look up by: their ids.wikipedia is a link, not a key. Keywords, fields, subfields and domains are looked up by their OpenAlex ID only.

For authors, lookup by ORCID resolves by the primary ORCID or any observed ORCID, not just the one currently shown as orcid.

When several records share an ID, the lookup returns one: for an ORCID, the profile that holds it as its primary orcid (over one that only observed it), then the one with the most works; for every other ID, the record with the most works (a funder over a sub-unit that carries the same ROR, a publishing group over an imprint). To see all of them, use the matching filter, e.g. /authors?filter=orcid:0000-0002-5787-4927.

Component ids: locations

Locations have no letter-prefixed OpenAlex ID. Their id is the identifier the copy was harvested under, prefixed with its namespace (doi:, pmh:, pmid:, mag:). GET /locations/{id} accepts the bare id, the locations/-prefixed key, or the full openalex.org URL:

curl "https://api.openalex.org/locations/doi:10.7717/peerj.4375"
curl "https://api.openalex.org/locations/locations/doi:10.7717/peerj.4375"
curl "https://api.openalex.org/locations/https://openalex.org/locations/doi:10.7717/peerj.4375"

Unlike OpenAlex IDs these are case-sensitive and contain / and :; copy them verbatim from a response. See Locations.

Merged entities

When duplicate works, authors, sources, institutions or funders are merged, requesting the old ID usually returns the record it was merged into. The API redirects the old ID to the surviving one and follows that redirect for you, so you get a normal 200 response whose id is the surviving ID:

$ curl -s "https://api.openalex.org/authors/A5107411446?select=id,display_name"
{"id": "https://openalex.org/A5107863564", "display_name": "Sonia Luquín"}

To tell that an ID was merged, compare the id in the response with the ID you asked for. Filters resolve the same way: filter=authorships.author.id:A5107411446 returns the surviving author’s works, and filter=cites:<merged work ID> returns the works citing the surviving work. Merges are followed to the end of the chain (if A was merged into B and B later into C, A returns C).

When a filter names merged IDs, the list response says what each one was read as, in meta.merged_ids. That’s how a batched lookup maps each record back to the ID you asked for:

$ curl -s "https://api.openalex.org/authors?filter=openalex_id:A5107411446|A5023888391&select=id,display_name"
{"meta": {"count": 2, ..., "merged_ids": [{"id": "https://openalex.org/A5107411446", "merge_into_id": "https://openalex.org/A5107863564"}]},
 "results": [{"id": "https://openalex.org/A5107863564", "display_name": "Sonia Luquín"}, {"id": "https://openalex.org/A5023888391", ...}]}

Each entry has the same two names as the snapshot’s merged_ids.csv.gz: id is the merged-away ID from your filter and merge_into_id is the record it resolves to. An ID that wasn’t merged isn’t listed: it comes back as itself. The field is there only when a filter named a merged-away work, author, source, institution or funder ID.

Redirecting merged IDs is a best-effort convenience, not a permanence promise: we do it while it’s practical, and we may change or stop it for some IDs or entity types. If you need a link that lasts, store the work’s DOI (or the author’s ORCID) alongside its OpenAlex ID.

A deleted ID (one with no surviving record) returns 404. For bulk repair of stored IDs, the snapshot publishes every merged-away work, author, source and institution ID with its surviving ID in merged_ids.csv.gz.

Last updated

View as Markdown