The Crossref REST API has been updated to use Elasticsearch instead of Solr -- read more about that project here.

This documentation is deprecated. Documentation for the new API can still be found at https://api.crossref.org/, and issues should be filed to the Crossref issue repository in Gitlab.

• Crossref REST API
  • Preamble
  • Meta
  • API overview
  • Result types
  • Resource components
  • Parameters
  • Queries
  • Field Queries
  • Sorting
  • Facet counts
  • Filter names
  • Result controls
  • API versioning
  • Documentation history

The Crossref REST API is one of a variety of tools and APIs that allow anybody to search and reuse our members' metadata in sophisticated ways.

If you read nothing else, please at least look at the API TIPs document and the "Etiquette" section of this document. It will save you (and us) much heartburn.

Records typically appear in the REST API within 20 minutes of their having been successfully deposited with Crossref.

Summary information (e.g. counts, etc.) are processed in batch every 24 hours.

We record and report service issues on our status page.

You might want to check this to see if we are already aware of a problem before you report it.

We also post notice of any ongoing performance problems with our services on our twitter feeds at CrossrefOrg and CrossrefSupport.

Report performance/availability at our support site.

Please report bugs with the API or the documentation on our issue tracker.

This work is licensed under a Creative Commons Attribution 4.0 International License.

Crossref asserts no claims of ownership to individual items of bibliographic metadata and associated Digital Object Identifiers (DOIs) acquired through the use of the Crossref Free Services. Individual items of bibliographic metadata and associated DOIs may be cached and incorporated into the user's content and systems.

We also have a privacy policy.

You might be able to avoid reading all this documentation if you instead use one of the several excellent libraries that have been written for the Crossref REST API. For example:

• crossref-commons (Python, developed by Crossref)
• habanero (Python)
• serrano (Ruby)
• rcrossref (R)
• crossrefapi (Python)
• crossref-rs (rust)
• pitaya (Julia)

If you know of another library you would like to see listed here, please let us know about it via the issue tracker.

We want to provide a public, open, and free API for all. And we don't want to unnecessarily burden developers (or ourselves) with cumbersome API tokens or registration processes in order to use the public REST API. For that to work, we ask that you be polite and try not to do anything that will take the public REST API down or otherwise make it unusable for others. Specifically, we encourage the following polite behaviour:

• Cache data so you don't request the same data over and over again.
• Actively monitor API response times. If they start to go up, back-off for a while. For example, add pauses between requests and/or reduce the number of parallel requests.
• Specify a User-Agent header that properly identifies your script or tool and that provides a means of contacting you via email using "mailto:". For example: GroovyBib/1.1 (https://example.org/GroovyBib/; mailto:[email protected]) BasedOnFunkyLib/1.4.

This way we can contact you if we see a problem.

• report problems and/or ask questions on our issue tracker.

Alas, not all people are polite. And for this reason we reserve the right to impose rate limits and/or to block clients that are disrupting the public service.

But we prefer carrots to sticks. As of September 18th 2017 any API queries that use HTTPS and have appropriate contact information will be directed to a special pool of API machines that are reserved for polite users.

Why are we doing this? Well- we don't want to force users to have to register with us. But this means that if some user of the public server writes a buggy script or ignores timeouts and errors- they can really bring the API service to its knees. What's more, it is very hard for us to identify these problem users because they tend to work off multiple parallel machines and use generic User-Agent headers. They are effectively anonymous. We're starting to have to spend a lot of time dealing with these problems and the degraded performance of the public API is affecting all the polite users as well.

So... we are keeping the public service as is. It will probably continue to fluctuate widely in performance. But now, if a client connects to the API using HTTPS and provides contact information either in their User-Agent header or as a parameter on their queries, then we will send them to a separate pool of machines. We expect to be able to better control the performance of these machines because, if a script starts causing problems, we can contact the people responsible for the script to ask them to fix it. Or, in extremis, we can block it.

How does it work? Simple. You can do one of two things to get directed to the "polite pool":

1. Include a "mailto" parameter in your query. For example:

https://api.crossref.org/works?filter=has-full-text:true&[email protected]

2. Include a "mailto:" in your User-Agent header. For example:

GroovyBib/1.1 (https://example.org/GroovyBib/; mailto:[email protected]) BasedOnFunkyLib/1.4.

Note that this only works if you query the API using HTTPS. You really should be doing that anyway (wags finger).

Q: Will you spam me with marketing bumf once you have our contact info?

A: No. We will only use it to contact you about problems with your scripts.

Q: Is this a secret plot to kill public access to your API?

A: No. It is an attempt to keep the public API reliable.

Q: What if I provide fake or incorrect contact info?

A: That is not very polite. If there is a problem and you don't respond, we'll block you.

Q: Does the contact info have to be a real name?

A: No. As long as somebody actually receives and pays attention to email at the address, it can be pseudo-anonymous, or whatever.

From time to time Crossref needs to impose rate limits to ensure that the free API is usable by all. Any rate limits that are in effect will be advertised in the X-Rate-Limit-Limit and X-Rate-Limit-Interval HTTP headers.

For ease-of-parsing, the X-Rate-Limit-Interval will always be expressed in seconds. So, for example the following tells you that you should expect to be able to perform 50 requests a second:

X-Rate-Limit-Limit: 50
X-Rate-Limit-Interval: 1s

Note that if we wanted to adjust the measurement window, we could specify:

X-Rate-Limit-Limit: 3000
X-Rate-Limit-Interval: 60s

This is always our last resort, and you can generally avoid it if you include contact information in the User-Agent header or mailto parameter as described above.

But seriously... this is a bummer. We really want you to use the API. If you are polite about it, you shouldn't have any problems.

What if you want to use our API for a production service that cannot depend on the performance uncertainties of the free and open public API? What if you don't want to be affected by impolite people who do not follow the API Etiquette guidelines? Well, if you’re interested in using these tools or APIs for production services, we have a service-level offering called "Plus". This service provides you with access to all supported APIs and metadata, but with extra service and support guarantees.

When you sign up for the Plus service, you will be issued an API token that you should put in the Authorization header of all your rest API requests. This token will ensure that said requests get directed to a pool of machines that are reserved for "Plus" SLA users. For example, with curl:

curl -X GET \
  https://api.crossref.org/works \
  -H 'Crossref-Plus-API-Token: Bearer yJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJpc3MiOiJodHRwOi8vY3Jvc3NyZWYub3JnLyIsImF1ZXYZImVuaGFuY2VkY21zIiwianRpIjoiN0M5ODlFNTItMTFEQS00QkY3LUJCRUUtODFCMUM3QzE0OTZEIn0.NYe3-O066sce9R1fjMzNEvP88VqSEaYdBY622FDiG8Uq' \
  -H 'User-Agent: GroovyBib/1.1 (https://example.org/GroovyBib/; mailto:[email protected]) BasedOnFunkyLib/1.4'

Note that you can still be "polite" and identify yourself as well. And, of course, replace the fake token above with the real token.

The API is generally RESTFUL and returns results in JSON. JSON formats returned by the API are documented here.

The API supports HTTP and HTTPS. Examples here are provided using HTTPS.

You should always url-encode DOIs and parameter values when using the API. DOIs are notorious for including characters that break URLs (e.g. semicolons, hashes, slashes, ampersands, question marks, etc.).

Note url-encoding applies to cursors as well since they too are likely to include characters that will break URLs (e.g. + symbol)

Note that, for the sake of clarity, the examples in this document do not url-encode DOIs or parameter values.

The API will only work for Crossref DOIs. You can test the registration agency for a DOI using the following route:

https://api.crossref.org/works/{doi}/agency

Testing the following Crossref DOI:

10.1037/0003-066X.59.1.29

Using the URL:

https://api.crossref.org/works/10.1037/0003-066X.59.1.29/agency

Will return the following result:

{
  status: "ok",
  message-type: "work-agency",
  message-version: "1.0.0",
  message: {
    DOI: "10.1037/0003-066x.59.1.29",
    agency: {
      id: "crossref",
      label: "Crossref"
    }
  }
}

If you use any of the API calls listed below with a non-Crossref DOI, you will get a 404 HTTP status response. Typical agency IDs include crossref, datacite, medra and also public for test DOIs.

All results are returned in JSON. There are three general types of results:

• Singletons
• Headers-only
• Lists

The mime-type for API results is application/vnd.crossref-api-message+json

Singletons are single results. Retrieving metadata for a specific identifier (e.g. DOI, ISSN, funder_identifier) typically returns in a singleton result.

You can use HTTP HEAD requests to quickly determine "existence" of a singleton. The advantage of this technique is that it is very fast because it does not return any metadata- it only returns headers and an HTTP status code (200=exists, 404=does not exist).

To determine if member ID 98 exists:

curl --head "https://api.crossref.org/members/98"

To determine if a journal with ISSN 1549-7712 exists:

curl --head "https://api.crossref.org/journals/1549-7712"

Lists results can contain multiple entries. Searching or filtering typically returns a list result. A list has two parts:

• Summary, which include the following information:

  • status (e.g. "ok", error)
  • message-type (e.g. "work-list" )
  • message-version (e.g. 1.0.0 )

• Items, which will will contain the items matching the query or filter.

Note that the "message-type" returned will differ from the mime-type:

• funder (singleton)
• prefix (singleton)
• member (singleton)
• work (singleton)
• work-list (list)
• funder-list (list)
• prefix-list (list)
• member-list (list)

Normally, an API list result will return both the summary and the items. If you want to just retrieve the summary, you can do so by specifying that the number of rows returned should be zero.

If the API call includes a query, then the sort order will be by the relevance score. If no query is included, then the sort order will be by DOI update date.

Crossref metadata records can be quite large. Sometimes you just want a few elements from the schema. You can "select" a subset of elements to return using the select parameter. This can make your API calls much more efficient. For example:

https://api.crossref.org/works?sample=10&select=DOI,title

Major resource components supported by the Crossref API are:

• works
• funders
• members
• prefixes
• types
• journals

These can be used alone like this

Resource components can be used in conjunction with identifiers to retrieve the metadata for that identifier.

The works component can be appended to other resources.

Parameters can be used to query, filter and control the results returned by the Crossref API. They can be passed as normal URI parameters or as JSON in the body of the request.

Multiple filters can be specified by separating name:value pairs with a comma:

https://api.crossref.org/works?filter=has-orcid:true,from-pub-date:2004-04-04

https://api.crossref.org/funders/100000015/works?query=global+state&filter=has-orcid:true&rows=1

Free form search queries can be made, for example, works that include renear and ontologies:

https://api.crossref.org/works?query=renear+ontologies

Field queries are available on the /works route and allow for queries that match only particular fields of metadata. For example, this query matches records that contain the tokens richard or feynman (or both) in any author field:

https://api.crossref.org/works?query.author=richard+feynman

Field queries can be combined with the general query paramter and each other. Each query parameter is ANDed with the others:

https://api.crossref.org/works?query.bibliographic=room+at+the+bottom&query.author=richard+feynman

These field queries are available on the /works route:

Note that query.title has been deprecated and is no longer available. Please use query.bibliographic instead.

Results from a listy response can be sorted by applying the sort and order parameters. Order sets the result ordering, either asc or desc. Sort sets the field by which results will be sorted. Possible values are:

An example that sorts results in order of publication, beginning with the least recent:

https://api.crossref.org/works?query=josiah+carberry&sort=published&order=asc

Facet counts can be retrieved by enabling faceting. Facets are enabled by providing facet field names along with a maximum number of returned term values. The larger the number of returned values, the longer the query will take. Some facet fields can accept a * as their maximum, which indicates that all values should be returned.

Facets are specified with the facet parameter:

https://api.crossref.org/works?rows=0&facet=type-name:*

Filters allow you to narrow queries. All filter results are lists.

The following filters are supported for the /works route:

The following filters are supported for the /members route:

The following filters are supported for the /funders route:

Multiple filters can be specified in a single query. In such a case, different filters will be applied with AND semantics, while specifying the same filter multiple times will result in OR semantics - that is, specifying the filters:

• is-update:true
• from-pub-date:2014-03-03
• funder:10.13039/100000001
• funder:10.13039/100000050

would locate documents that are updates, were published on or after 3rd March 2014 and were funded by either the National Science Foundation (10.13039/100000001) or the National Heart, Lung, and Blood Institute (10.13039/100000050). These filters would be specified by joining each filter together with a comma:

/works?filter=is-update:true,from-pub-date:2014-03-03,funder:10.13039/100000001,funder:10.13039/100000050

A filter with a dot in its name is special. The dot signifies that the filter will be applied to some other record type that is related to primary resource record type. For example, with work queries, one can filter on works that have an award, where the same award has a particular award number and award-giving funding agency:

/works?filter=award.number:CBET-0756451,award.funder:10.13039/100000001

Here we filter on works that have an award by the National Science Foundation that also has the award number CBET-0756451.

The prefix of a Crossref DOI does NOT indicate who currently owns the DOI. It only reflects who originally registered the DOI. Crossref metadata has an owner_prefix element that records the current owner of the Crossref DOI in question.

Crossref also has member IDs for depositing organisations. A single member may control multiple owner prefixes, which in turn may control a number of DOIs. When looking at works published by a certain organisaton, member IDs and the member routes should be used.

Note that dates in filters should always be of the form YYYY-MM-DD, YYYY-MM or YYYY. Also note that date information in Crossref metadata can often be incomplete. So, for example, a publisher may only include the year and month of publication for a journal article. For a monograph they might just include the year. In these cases the API selects the earliest possible date given the information provided. So, for instance, if the publisher only provided 2013-02 as the published date, then the date would be treated as 2013-02-01. Similarly, if the publisher only provided the year 2013 as the date, it would be treated at 2013-01-01.

When using time filters to retrieve periodic, incremental metadata updates, the from-index-date filter should be used over from-update-date, from-deposit-date, from-created-date and from-pub-date. The timestamp that from-index-date filters on is guaranteed to be updated every time there is a change to metadata requiring a reindex.

You can control the delivery and selection results using the rows, offset and sample parameters.

If you are expecting results beyond 10K, then use a cursor to deep page through the results. Note that not all routes support cursors.

Normally, results are returned 20 at a time. You can control the number of results returns by using the rows parameter. To limit results to 5, for example, you could do the following:

https://api.crossref.org/works?query=allen+renear&rows=5

If you would just like to get the summary of the results, you can set the rows to 0 (zero).

https://api.crossref.org/works?query=allen+renear&rows=0

The maximum number rows you can ask for in one query is 1000.

The number of returned items is controlled by the rows parameter, but you can select the offset into the result list by using the offset parameter. So, for example, to select the second set of 5 results (i.e. results 6 through 10), you would do the following:

https://api.crossref.org/works?query=allen+renear&rows=5&offset=5

Offsets for /works are limited to 10K. Use cursor (see below) for larger /works results sets.

Using large offset values can result in extremely long response times. Offsets in the 100,000s and beyond will likely cause a timeout before the API is able to respond. An alternative to paging through very large result sets (like a corpus used for text and data mining) is to use the API's exposure of Solr's deep paging cursors. Any combination of query, filters and facets may be used with deep paging cursors. While rows may be specified along with cursor, offset and sample cannot be used. To use deep paging make a query as normal, but include the cursor parameter with a value of *. In this example we will page through all journal-article works from member 311:

https://api.crossref.org/members/311/works?filter=type:journal-article&cursor=*

A next-cursor field will be provided in the JSON response. To get the next page of results, pass the value of next-cursor as the cursor parameter. For example:

https://api.crossref.org/members/311/works?filter=type:journal-article&cursor=AoE/CGh0dHA6Ly9keC5kb2kub3JnLzEwLjEwMDIvdGRtX2xpY2Vuc2VfMQ==

Note that the actual cursor value will be different from this illustration.

Clients should check the number of returned items. If the number of returned items is fewer than the number of expected rows then the end of the result set has been reached. Using next-cursor beyond this point will result in responses with an empty items list.

The cursor parameter is available on all /works resources.

Being able to select random results is useful for both testing and sampling. You can use the sample parameter to retrieve random results. So, for example, the following select 10 random works:

https://api.crossref.org/works?sample=10

Note that when you use the sample parameter, the rows and offset parameters are ignored.

All works published by owner prefix 10.1016 in January 2010

https://api.crossref.org/prefixes/10.1016/works?filter=from-pub-date:2010-01,until-pub-date:2010-01

All works funded by 10.13039/100000001 that have a CC-BY license

https://api.crossref.org/funders/10.13039/100000001/works?filter=license.url:https://creativecommons.org/licenses/by/3.0/

All works published by owner prefix 10.6064 from February 2010 to February 2013 that have a CC-BY license

https://api.crossref.org/prefixes/10.6064/works?filter=license.url:https://creativecommons.org/licenses/by/3.0/,from-pub-date:2010-02,until-pub-date:2013-02

All works funded by 10.13039/100000015 where license = CC-BY and embargo <= 365 days

https://api.crossref.org/funders/10.13039/100000015/works?filter=license.url:https://creativecommons.org/licenses/by/3.0/,license.delay:365

Note that the filters for license URL and maximum license embargo period (license.url and license.delay) combine to filter each document's metadata for a license with both of these properties.

All works where the archive partner listed = 'CLOCKSS'

https://api.crossref.org/works?filter=archive:CLOCKSS

All members with hind in their name (e.g. Hindawi)

https://api.crossref.org/members?query=hind

All licenses linked to works published by Elsevier

https://api.crossref.org/v1/works?facet=license:*&filter=member:78&rows=0

All licenses applied to works published in the journal Pathology Research International

https://api.crossref.org/works?facet=license:*&filter=issn:2090-8091

All works with an award numbered roughly 1 F31 MH11745 also awarded by funder with ID 10.13039/100000025

https://api.crossref.org/works?filter=award.number:1F31MH11745,award.funder:10.13039/100000025

The number of DOIs that have references AND where references are open faceted by publisher name

https://api.crossref.org/v1.0/works?filter=has-references:true,reference-visibility:open&facet=publisher-name:*&rows=0

In theory, the syntax of the API can vary independently of the result representations. In practice, major version changes in either will require changes to API clients and so versioning of the API will apply to both the API syntax and the result representation.

The API uses a semantic versioning scheme whereby the version number is divided into three parts delimited by periods. The first number represents the "major" release number. The second represents a "minor" release number.

Version 1.20
        ^  ^
        |  |
    major  |
       minor

Major version increments are defined as releases that can break backwards compatibility. Crossref will only commit to supporting the latest two major releases simultaneously and legacy major releases will be supported for no more than nine months. Exceptions to these rules may be made when major releases are required to ensure the security or stability of the system.

Minor version increments are defined as backwards compatible. There is no limit on the number of minor versions that Crossref can roll out. Note that client applications should not have dependencies on minor versions, and Crossref will only maintain the latest minor version for the two most recent major versions.

Adding syntax options or metadata to representations will normally be backwards compatible and will thus normally only trigger minor version changes. Renaming or restructuring syntax options of metadata tends not to be backward compatible and will thus typically trigger major version changes

If you need to tie your implementation to a specific major version of the API, you can do so by using version-specific routes. The default route redirects to the most recent version of the API. Some older major versions may be available using a version prefix. For example, to access version v1 of the API:

https://api.crossref.org/v1/works

Each major version has no backwards incompatible changes within its public interface.

• V1: 2013-09-08, first draft.
• V2: 2013-09-24, reference platform deployed
• v3: 2013-09-25, reworked filters. Added API versioning doc
• v4: 2013-09-25, more filter changes.
• v5: 2013-09-27, doc mime-type and message-type relationship
• v6: 2013-10-01, updated sample & added examples with filters
• v6: 2013-10-01, corrected warning date
• v7: 2013-10-02, fixed typos
• v8: 2013-10-17, updated warning. Added email address
• v9: 2013-12-13, update example urls
• v10: 2013-12-13, /types routes, type filter, issn filter
• v11: 2013-12-14, indexed timestamps, has-archive and archive implemented
• v12: 2014-01-06, directory filter
• v13: 2014-02-10, new /members, /publishers becomes /prefixes, new member filter, publisher filter becomes prefix
• v14: 2014-02-14, new has-funder filter.
• v15: 2014-02-27, new /licenses route
• v16: 2014-05-19, new /journals route, new CrossMark (updates and update policy) filters, new sort and order parameters
• v17: 2014-05-19, new facet query parameter
• v18: 2014-05-29, new /works/{doi}/agency route
• v19: 2014-06-23, new textual filters - container-title, category-name.
• v20: 2014-06-24, OR filter queries, type-name filter.
• v21: 2014-07-01, new award.number and award.funder relational filters.
• v22: 2014-07-16, changed title to more accurately reflect scope of API.
• v23, 2014-09-01, semantics of mutliple filters, dot filters
• v24, 2014-10-15, added info on license of Crossref metadata itself. Doh.
• v25, 2015-05-06, added link to issue tracker. Removed Warning section.
• v26, 2015-10-20, added new filters - from-created-date, until-created-date, affiliation, has-affiliation, assertion-group, assertion, article-number, alternative-id
• v27, 2015-10-30, added cursor parameter to /works resources
• v28, 2016-05-09, added link to source of category lables
• v29, 2016-05-24, added field queries
• v30, 2016-09-26, highlight issue tracker
• v31, 2016-10-05, document has-clinical-trial-number and has-abstract filters
• v32, 2016-10-27, document rate limit headers
• v33, 2016-11-07, guidance on when to use offset vs cursor
• v34, 2017-04-26, document support for HTTPS. Update examples to use HTTPS.
• v35, 2017-04-26, document use of head reqeusts to determine existence
• v36, 2017-04-27, fixed license route examples to use facet/filter instead
• v37, 2017-04-27, query.bibliographic
• v38, 2017-04-27, add v1.1 filters and sort fields
• v39, 2017-04-27, remove mention of dismax
• v40, 2017-04-27, clarify faceting feature
• v41, 2017-04-28, document sample max = 100, clarify cursors only work on some routes
• v42, 2017-04-28, life, the universe, and everything
• v43, 2017-04-28, reminder on the wisdom of url-encoding
• v44, 2017-04-28, clarify that field queries apply to /works route
• v45, 2017-04-28, document location filter for /funders route
• v46, 2017-06-14, minor text changes and new funder registry link
• v47, 2017-07-04, clarify query.affiliation
• v48, 2017-07-13, correct "first and given" names to "given and family"
• v49, 2017-07-20, move document version history, add section on libraries
• v50, 2017-07-20, add TOC, move document history, add etiquet section, add production use section, general formatting + cleanup
• v51, 2017-07-24, clarified license of the documentation (as opposed to metadata)
• v52, 2017-07-27, removed service notice and what's new section.
• v53, 2017-08-11, mention full-text.application filter
• v54, 2017-09-18, add info about new "polite pool"
• v55, 2017-09-21, document /member and /funder filters. document publisher-name facet. document select parameter.
• v56, 2018-01-26, add info on frequency of indexing
• v57, 2018-02-01, document ISBN filter
• v58, 2018-02-13, document reference-visibility filter for /works and /members routes
• v59, 2018-02-13, added info about Mtedata Plus service. Corrected spelling. Added example of using reference-visibility filter.
• v60, 2018-02-22, added info for "Plus" users on use of token in Authorization header.
• v61, 2018-02-26, add curl example for use of token.
• v62, 2018-06-18, clarify how to parse X-Rate-Limit-Limit-Interval
• v63, 2018-08-16, remove mistakenly listed year facet. published is correct facet name.
• v64, 2018-09-04, add text and link to status page.
• v65, 2019-07-17, updated Plus token name to new recommendation
• v66, 2019-10-16, deprecated query.title for query.bibliographic
• v67, 2020-04-28, clarify that url-encoding should be applied to all parameters including cursors.
• v68, 2021-02-22, Added link to Pitaya Julia library
• v69, 2021-04-30, Update preamble with reading advise