OpenCitations DOI Citation Graph Extractor
Pricing
from $0.49 / 1,000 results
OpenCitations DOI Citation Graph Extractor
Look up a scholarly DOI in the OpenCitations index to get every paper that cites it, every paper it references, or its total citation count, with cross-walked DOI, OMID, PMID and OpenAlex identifiers.
Pricing
from $0.49 / 1,000 results
Rating
0.0
(0)
Developer
Dataspryva
Maintained by CommunityActor stats
0
Bookmarked
2
Total users
1
Monthly active users
7 days ago
Last modified
Categories
Features
Share
OpenCitations publishes a fully open citation graph of scholarly works. Give this actor one or more DOIs — or PMIDs, OMIDs, ISSNs, ISBNs — and it returns the graph around them: the papers that cite each work, the papers each work cites, both directions in one run, citation and reference totals, single citation edges, and full bibliographic metadata. Everything comes from OpenCitations' public Index and Meta APIs, so no key, login, cookie or proxy is needed.
Modes
| Mode | What it returns |
|---|---|
getCitations | One row per work that cites each identifier (incoming citations) |
getReferences | One row per work each identifier cites (outgoing references) |
getGraph | Citations and references together, split evenly by maxItems |
getCitationCount | Incoming citation total for each identifier |
getReferenceCount | Outgoing reference total for each identifier |
getCounts | Citation and reference totals in one record (plus venue count for an ISSN) |
getVenueCitationCount | All incoming citations to a journal identified by ISSN |
getMetadata | Title, authors, venue, publisher, volume, issue, pages, type and dates |
getCitation | A single citation edge looked up by its OCI |
Inputs
| Field | Type | Default | Notes |
|---|---|---|---|
mode | select | getCitations | One of the modes above |
identifiers | string list | [] | DOIs, DOI URLs, doi: values, PMIDs, OMIDs, ISSNs or ISBNs; one per line |
oci | string | empty | Only for getCitation, e.g. 06101801781-06180334099 |
maxItems | integer | 100 (prefill 3) | Row cap for edge modes; identifier cap for count and metadata modes |
startPage | integer | 1 | Resume or slice a large edge result set; each page holds up to 1,000 edges |
includeWorkMetadata | boolean | false | Add citingTitle, citingAuthors, citingPubDate, citingVenue and the cited equivalents to edge rows |
Saved tasks that still send the old doi key keep working: the actor maps it onto identifiers in code. The old mode, maxItems and output fields are unchanged.
Examples
Who cites a paper:
{"mode": "getCitations","identifiers": ["10.1186/1756-8722-6-59"],"maxItems": 100}
Both directions for two papers:
{"mode": "getGraph","identifiers": ["https://doi.org/10.1038/leu.2016.153","10.7717/peerj-cs.421"],"maxItems": 200}
Counts and metadata:
{"mode": "getCounts","identifiers": ["10.1186/1756-8722-6-59", "issn:0138-9130"]}
{"mode": "getMetadata","identifiers": ["pmid:23958373", "10.7717/peerj-cs.421"]}
A single edge by OCI:
{"mode": "getCitation","oci": "06101801781-06180334099"}
Output
Records are flat and omit-empty: a field appears only when OpenCitations has a real value for it.
Citation and reference edge
oci— Open Citation Identifier for the edgequeryIdentifier— the identifier you asked forresolvedIdentifier— the DOI/PMID/OMID used after resolving an OpenAlex or PMCID inputciting,cited— raw identifier token stringscitingDoi,citedDoi— DOI of each sidecitingOmid,citedOmid— OpenCitations Meta ID of each sidecitingPmid,citedPmid— PubMed ID of each side, when knowncitingOpenalex,citedOpenalex— OpenAlex ID of each side, when knowncitingUrl,citedUrl— DOI resolver linkscitingPubmedUrl,citedPubmedUrl— PubMed links, when a PMID is knowncreation— publication date of the citing worktimespan— ISO-8601 duration between the cited and citing publication datesjournalSc,authorSc— journal / author self-citation flagsjournalScTuple— the two flags togethercitingTitle,citingAuthors,citingPubDate,citingVenue— only withincludeWorkMetadatacitedTitle,citedAuthors,citedPubDate,citedVenue— only withincludeWorkMetadatasourceUrl— exact OpenCitations page URLrecordType—citationorreferencescrapedAt— ISO timestamp
Counts
identifier, plusdoi/issn/pmid/omidas applicablecitationCount— incoming citationsreferenceCount— outgoing references (getCounts,getReferenceCount)venueCitationCount— citations to an ISSN (getCountswhen the identifier is an ISSN,getVenueCitationCount)sourceUrl— API URL usedrecordType—citationCount,referenceCount,venueCitationCountorcountsscrapedAt
Metadata
identifier,doi,omid,pmid,pmcid,openalexId,issn,issns,isbntitle,author,authors(name + ORCID/OMID when known)pubDate,publicationYearvolume,issue,page,publicationTypevenue,venueName,venueIssn,venueOmid,venueOpenalexpublisher,publisherName,publisherCrossref,publisherOmideditor,editorsdoiUrl,sourceUrlrecordType—metadatascrapedAt
Error row
Pushed instead of a normal record for an invalid identifier, an unresolved identifier, an empty lookup, or an API failure.
error— short code such asinvalid_identifier,invalid_oci,missing_identifier,unresolved_identifier,not_foundorfetch_failedmessage— plain-language explanationinput— the value that caused itmode— the mode being runsourceUrlrecordType—errorscrapedAt
Pagination and large graphs
Citation and reference modes use OpenCitations' server-side page / page_size parameters. The actor keeps requesting pages until maxItems is reached or the source runs out, deduplicates edges by OCI, and never treats the natural end of results as a failure. startPage lets you resume a very large graph in slices; maxItems is the total row budget across all identifiers in the run.
Data source and scope
- Citations, references, counts and OCIs: OpenCitations Index API v2 (
api.opencitations.net/index/v2). - Titles, authors, venues, publishers and dates: OpenCitations Meta API v1 (
api.opencitations.net/meta/v1). - OpenCitations rate-limits anonymous API use to 180 requests per minute per IP. The actor paces its page requests; extremely large graphs can take a while.
- The Index contains citation links, not full texts or abstracts.
- OpenAccess/PMCID and OpenAlex identifiers are accepted directly by
getMetadata; in edge and count modes the actor resolves them to a DOI/OMID/PMID through Meta first. - Metadata is only as complete as the OpenCitations record: missing authors, venues or pages are omitted rather than guessed.
- This is an independent third-party actor. It is not operated by or affiliated with OpenCitations.
FAQ
Do I need an API key? No. The OpenCitations APIs used here are open and cookieless.
Can I look up several papers in one run? Yes — put one identifier per line in identifiers. maxItems is shared across all of them, and each row carries queryIdentifier so you can group the output.
What identifier formats are accepted? Bare DOIs, doi: values, https://doi.org/… URLs, PMIDs, omid:br/…, issn:…, isbn:… and, when OpenCitations has them, openalex:W… and pmcid:PMC… for metadata. Bare 8-digit values are treated as PMIDs; use issn:/isbn: prefixes when that is what you have.
Why is a citation row missing a PMID or OpenAlex ID? OpenCitations only stores the identifiers it has cross-walked for that work. Missing identifiers are omitted.
Why does getGraph return fewer citations than getCitations with the same maxItems? getGraph splits the row budget between directions, so both halves appear in one dataset. Run getCitations or getReferences for a full single-direction budget.
Can I get more than 100,000 rows? maxItems is capped at 100,000 per run. Use startPage to continue a larger result set in another run.
How fresh is the data? Every run queries OpenCitations live, so results reflect the current index.