DOI Metadata & Bibliography Enrichment

Paste a list of DOIs and get one clean citation record per DOI. This bulk DOI metadata lookup checks each DOI at doi.org, reads the metadata from the official Crossref, DataCite or JaLC API, and returns title, authors, journal or book, exact dates, volume, pages, ISSN/ISBN, license and citation counts.

Use it to fill a reference list, check a bibliography or enrich a spreadsheet of papers and datasets. The smallest run is one DOI, and three DOIs finish in about 2 seconds.

Table of contents

• What it does
• Quick start
• Use cases
• Input
• Output
• Pricing and limits
• API, export and integrations
• Troubleshooting
• FAQ
• Data sources and licence
• Support

What it does

A DOI (Digital Object Identifier) is a permanent ID for a paper, book chapter, dataset or other work, for example 10.1038/nature12373. Each DOI is registered with one registration agency. This Actor:

1. Cleans your input. It accepts 10.…, doi:10.…, https://doi.org/10.… and dx.doi.org links, in any letter case.
2. Asks doi.org which agency registered each DOI.
3. Reads the record from that agency: Crossref (most journals and books), DataCite (datasets, software, repositories) or JaLC (Japan).
4. Writes one Dataset row for each input value, in your input order, with a clear status.

Good fit

• Turn a list of DOIs into a table of titles, authors, journals, years and pages.
• Check a reference list: find DOIs that are not registered or have a typo.
• Add ISSN, ISBN, publisher, license and citation counts to a list of papers.
• Mix journal articles and research datasets in one run (Crossref and DataCite).

Not a fit / limits

• No search by title or author. You need the DOI.
• No PDFs, full text or abstracts. The Actor reads only bibliographic metadata.
• No Open Access links (for example from Unpaywall).
• DOIs from other agencies (for example mEDRA) are only identified. The row has status: unsupported_agency and the agency name, but no title or authors.
• No BibTeX or RIS file. You get structured fields that you can export as JSON, CSV or Excel.
• Up to 500 DOIs per run.

Quick start

1. Open the Input tab.
2. Keep the three example DOIs in DOIs (dois) or paste your own, one per line.
3. Leave Contact email empty, or add your own email for large runs.
4. Click Start.
5. Open the Output tab, view Citations. Check status and title in the first row.

Expected result: one row per DOI. With the example input you get three complete rows: a Nature article (Crossref), a Dryad dataset (DataCite) and a book chapter (Crossref).

Use cases

Buyer job	Input path	Output entity	Key fields	Next integration
Build a reference table	dois	One citation record per DOI	title, authors, containerTitle, issuedDate, pages	Export CSV/Excel, or read the Dataset by API
Clean a bibliography	dois	Status per DOI	status, statusReason, errorCode, doi	Filter rows where status is not complete
Enrich a paper list	dois	Identifiers and counts	issn, isbn, publisher, licenses, citedByCount	Join on doi (always lower case)
Catalogue research data	dois	DataCite dataset record	resourceTypeGeneral, publisher, publicationYear, language	Export or API

Input

Input fields

Field	Plain name	Type	Required	Default / prefill	Allowed values	Effect	Example	Recommendation
dois	DOIs	array of strings (one per line in the form)	yes	Prefill: 3 example DOIs. No runtime default.	1–500 values, each up to 1,000 characters	Each value becomes one output row	10.1038/nature12373	Paste DOIs in any common form. Split lists longer than 500 into several runs.
contactEmail	Contact email for Crossref and DataCite	string	no	empty	A plain email address, up to 254 characters	Sent only in the request User-Agent (mailto) to Crossref, DataCite and doi.org. Crossref uses it to put requests in its "polite" pool.	name@example.org	Add it for runs of hundreds of DOIs. It is never logged or saved in the output.

How dois is read:

• Accepted forms: 10.1038/nature12373, doi:10.1038/nature12373, https://doi.org/10.1038/nature12373, http://dx.doi.org/… and percent-encoded links. Spaces around a value are removed. Empty lines are skipped.
• Letter case does not matter: DOIs are case-insensitive, and the output doi is always lower case.
• A repeated DOI (also in another form or case) gets its own row with the same data, but it is charged only once.
• A value that is not a DOI gets a free row with status: invalid.
• An empty list or more than 500 values is rejected before the run starts.

Input examples

Mixed list of papers and data

{
  "dois": [
    "10.1038/nature12373",
    "10.5061/dryad.8515",
    "10.1007/978-3-319-24277-4_9"
  ]
}

Expected behavior: doi.org returns Crossref for the first and third DOI and DataCite for the second. One batch request goes to each API. Expected output: three complete rows with title, authors, containerTitle, issuedDate and agency. Limit or cost note: 3 found DOIs = 3 doi-metadata events plus the run start.

Reference list check with problems

{
  "dois": [
    "https://doi.org/10.1038/NATURE12373",
    "10.9999/notexist",
    "not a doi",
    "doi:10.1038/nature12373"
  ],
  "contactEmail": "name@example.org"
}

Expected output: row 1 is complete; row 2 is not_found with errorCode: DOI_NOT_REGISTERED; row 3 is invalid (free); row 4 is a duplicate of row 1 (complete, free).

Output

Each input value gives exactly one row in the default Dataset, in input order. The Citations view shows the main columns; the Identifiers view shows IDs and links. A run summary is saved as RUN_SUMMARY in the default Key-value store.

Status values

status	Meaning	Charged as
complete	The DOI is registered and the record has a title.	doi-metadata
partial	The DOI is registered, but the record has no title (NO_TITLE, for example a journal issue or a figure) or the agency API has no record for it yet (RECORD_NOT_INDEXED). Read statusReason.	doi-metadata for NO_TITLE; doi-check for RECORD_NOT_INDEXED
not_found	doi.org says the DOI is not registered (DOI_NOT_REGISTERED).	doi-check
unsupported_agency	The DOI is registered with another agency. agency names it (UNSUPPORTED_AGENCY).	doi-check
invalid	The value is not a DOI (INVALID_DOI).	free
error	A source failed after retries, or the run reached its time limit. Existence is unknown; the row never says "not found" in this case.	free

Output fields

Field	Meaning	Type	Can be empty?	Example / enum	Source / notes
input	Your input value exactly as given	string	no	https://doi.org/10.1038/NATURE12373	Input
doi	DOI in lower case	string	null for invalid	10.1038/nature12373	Cleaned input
doiUrl	Resolvable link	string	null for invalid	https://doi.org/10.1038/nature12373	Built from doi
status	Result type	string	no	see Status values	Actor
statusReason	Plain-English reason for any status other than complete	string	null for complete	doi.org reports that this DOI does not exist …	Actor
agency	Registration agency	string	null for invalid, not_found, error	Crossref, DataCite, JaLC, mEDRA	doi.org
type	Work type as the agency names it	string	yes	journal-article, book-chapter, dataset	Agency
resourceTypeGeneral	DataCite general type	string	null outside DataCite	Dataset, Software, Text	DataCite
title	Title, markup removed, entities decoded	string	null for partial and non-found rows	Nanometre-scale thermometry in a living cell	Agency
subtitle	Subtitle	string	yes		Agency
authors	Authors in record order: given, family, name, orcid, affiliations	array	empty when the record has none	[{"given":"G.","family":"Kucsko",…}]	Agency. orcid only when the source lists it. affiliations is copied as the source has it and can include contact details.
editors	Editors, same shape	array	often empty		Agency
containerTitle	Journal, book or proceedings that contains the work	string	yes	Nature, ggplot2	Agency. For a book chapter: the book.
containerTitles	All container titles as the agency lists them	array	yes	["Use R!","ggplot2"]	Agency
seriesTitle	Book series	string	yes	Use R!	Crossref
publisher	Publisher	string	yes	Springer Science and Business Media LLC	Agency
issuedDate	Publication date at the source precision	string	yes	2013-07-31, 2013-08, 2016	Agency. Never padded to a full date.
publishedOnlineDate	Online date	string	yes	2013-07-31	Crossref
publishedPrintDate	Print date	string	yes	2013-08	Crossref
publicationYear	Year	integer	yes	2013	Agency
volume	Volume	string	yes	500	Agency
issue	Issue	string	yes	7460	Agency
pages	Page range	string	yes	54-58	Agency
articleNumber	Article number (journals without pages)	string	yes	e06495	Crossref
issn	ISSNs: value, type (print/electronic/null)	array	yes	[{"value":"0028-0836","type":"print"}]	Agency
isbn	ISBNs, same shape	array	yes	[{"value":"9783319242750","type":null}]	Agency
licenses	License statements: url, name, contentVersion, startDate	array	yes	[{"url":"http://www.springer.com/tdm",…}]	Crossref license / DataCite rightsList
referencesCount	References deposited with the record	integer	yes	30	Crossref
citedByCount	Citations known to the agency	integer	yes	1845	Crossref is-referenced-by-count / DataCite citationCount. Other indexes count differently.
language	Language code	string	null for Crossref	en, ja	DataCite, JaLC
landingUrl	Publisher landing page	string	yes	https://www.nature.com/articles/nature12373	Agency
sourceUpdatedAt	When the agency last indexed or updated the record	string	yes	2026-10-11T02:44:13Z	Agency
authorsCount	Number of authors	integer	no (0 when none)	8	Actor
provider	API the row came from	string	yes	Crossref REST API	Actor
sourceUrl	API request the metadata was read from	string	yes	https://api.crossref.org/works?filter=doi:…	Actor
fetchedAt	When the row was made (ISO 8601)	string	no	2026-10-11T09:31:37.688Z	Actor
errorCode	Machine code	string	null for complete	INVALID_DOI, DOI_NOT_REGISTERED, UNSUPPORTED_AGENCY, RECORD_NOT_INDEXED, NO_TITLE, SOURCE_UNAVAILABLE, RATE_LIMITED, DEADLINE_REACHED	Actor

null means the agency record does not have this value. The Actor never invents a value, for example a day or month that the source does not give.

Output entities and surfaces

Situation	Where	Result shape	User action
Normal item	Dataset, view Citations	complete row with metadata	Export or use the fields
DOI not registered	Dataset	not_found row, metadata fields null	Check the DOI for a typo
Partial result	Dataset	partial row with statusReason	Read statusReason; try again later for RECORD_NOT_INDEXED
Technical failure	Dataset + RUN_SUMMARY	error row with SOURCE_UNAVAILABLE, RATE_LIMITED or DEADLINE_REACHED	Run these DOIs again later; error rows are free
Run overview	Key-value store RUN_SUMMARY	counts by status and agency, HTTP stats, warnings, sources	Check status and counts

Output example

{
  "input": "10.1038/nature12373",
  "doi": "10.1038/nature12373",
  "doiUrl": "https://doi.org/10.1038/nature12373",
  "status": "complete",
  "statusReason": null,
  "agency": "Crossref",
  "type": "journal-article",
  "title": "Nanometre-scale thermometry in a living cell",
  "authors": [
    { "given": "G.", "family": "Kucsko", "name": "G. Kucsko", "orcid": null, "affiliations": [] }
  ],
  "containerTitle": "Nature",
  "publisher": "Springer Science and Business Media LLC",
  "issuedDate": "2013-07-31",
  "publishedPrintDate": "2013-08",
  "publicationYear": 2013,
  "volume": "500",
  "issue": "7460",
  "pages": "54-58",
  "issn": [
    { "value": "0028-0836", "type": "print" },
    { "value": "1476-4687", "type": "electronic" }
  ],
  "authorsCount": 8,
  "provider": "Crossref REST API",
  "errorCode": null
}

This example is trimmed: it shows one row from the first input example, with one of eight authors and some fields left out. Counts such as citedByCount change over time.

Field names compared with other DOI tools

If you move from another DOI tool, these names may help:

Common name elsewhere	Field here
journal	containerTitle
publishedDate	issuedDate
publishedYear	publicationYear
page	pages
citationCount	citedByCount
url	doiUrl (DOI link) or landingUrl (publisher page)
authors as one string	authors[].name

Pricing and limits

Pricing unit

This Actor uses pay-per-event pricing. You pay for each DOI result and once for each run start. Platform usage is included.

Event	When it is charged	Free plan	Bronze	Silver	Gold
doi-metadata	A complete row, or a partial row with metadata (NO_TITLE)	$0.0018	$0.0017	$0.0016	$0.0015
doi-check	A not_found, unsupported_agency or RECORD_NOT_INDEXED row	$0.0005	$0.0005	$0.0005	$0.0005
Actor start	Once per run (memory is 256 MB, so always one event)	$0.001	$0.001	$0.001	$0.001

Free: invalid rows, error rows and repeated DOIs. A run where no value is a DOI fails with a clear message and charges only the start.

Cost examples (Free plan prices)

Scenario	Input scope	Expected events	Base charge	Total
Example input	3 DOIs, all found	3 × doi-metadata	$0.001 start	$0.0064
Small list	100 DOIs, all found	100 × doi-metadata	$0.001 start	$0.181
Reference check	100 DOIs: 90 found, 5 not registered, 5 invalid	90 × doi-metadata, 5 × doi-check	$0.001 start	$0.1655
Full run	500 DOIs, all found	500 × doi-metadata	$0.001 start	$0.901

Spending limit

If you set a maximum cost per run, the Actor stops before it goes over the limit. Rows already saved stay in the Dataset; DOIs after the limit get no row, and RUN_SUMMARY.status is CAPPED.

Limits

• Up to 500 DOIs per run. In tests, 500 DOIs finished in about 24 seconds and 100 DOIs in about 6 seconds.
• The Actor reads sources for up to 240 seconds. DOIs not processed in time get a free error row with DEADLINE_REACHED.
• It sends batch requests (up to 50 DOIs each) and retries temporary failures. It respects source rate limits.

API, export and integrations

• API: start the Actor with the Apify API and read the Dataset items. The doi field is lower case, so you can join on it.
• Export: download the Dataset as JSON, CSV, Excel, XML or RSS from the Output tab.
• Scheduling: you can run it on an Apify schedule, for example to refresh citation counts.

Use with AI agents (MCP)

AI agents can call this Actor through the Apify MCP server: add https://mcp.apify.com?tools=cliqtomedia/doi-bibliography-enrichment to your MCP client (Claude, Cursor, VS Code and others). The MCP server needs your Apify account (OAuth sign-in or an Apify API token); the DOI sources need no key. The agent builds the input from the input schema. The smallest useful input:

{ "dois": ["10.1038/nature12373"] }

Key output fields for an agent: status (complete, not_found, …), title, authors, containerTitle, issuedDate, agency, errorCode. A call with 10 DOIs usually finishes in a few seconds and costs about $0.019. Up to 500 DOIs fit in one run. Treat error rows as "unknown", not as "does not exist".

Troubleshooting

Symptom	Check	Action
not_found for a DOI you trust	input, doi	Look for a typo or extra characters. doi.org itself says the DOI is not registered.
invalid row	input, statusReason	After cleaning, the value is not in the form 10.<prefix>/<suffix>. Fix the value.
partial with RECORD_NOT_INDEXED	statusReason	The DOI is registered, but the agency API does not have the record yet. Try again later.
partial with NO_TITLE	type	Records such as journal issues or figures often have no title. This is source data.
unsupported_agency	agency	The DOI belongs to another agency. This Actor does not read its metadata.
error rows	errorCode, RUN_SUMMARY.warnings	A source was down or slow. Run only these DOIs again later; they were not charged.
Run rejected at start	Input	dois is empty or has more than 500 values.

FAQ

Do I need an API key?

No. Crossref, DataCite, JaLC and doi.org are open. You only need your Apify account.

Why is a date only a year or a month?

The Actor keeps the precision of the source. If the record says 2016, you get 2016, not 2016-01-01.

Why is citedByCount different from Google Scholar?

It is the count known to the registration agency (Crossref or DataCite). Other indexes count different sources.

How fresh is the data?

Each run reads the sources live. Each row has fetchedAt (our read time) and sourceUpdatedAt (when the agency last updated the record).

Can I get abstracts or PDFs?

No. The Actor returns bibliographic metadata only.

Data sources and licence

• doi.org — registration agency of each DOI.
• Crossref REST API — bibliographic metadata. Abstracts are not exported.
• DataCite REST API — metadata under CC0.
• JaLC — Japanese records, read through doi.org content negotiation.

This Actor is independent and not endorsed by Crossref, DataCite, JaLC or the DOI Foundation.

Support

Use the Issues tab of this Actor. Add the run ID, the DOI and the status or errorCode you saw. Do not send API tokens or private data.