This is the reference documentation for the Zendesk Help Center API.
Any general mechanisms and conventions, such as pagination and authentication, work as described in the Zendesk API introduction and are considered implied in this text.
Responses are always filtered according to the permissions of the user on whose behalf the API request is made. For example, when the user requests a list of articles in a section, the API returns only the articles the user can view in Help Center.
The API works only with Help Center items. For example, comments can be made on both articles and requests, but the Help Center API only provides access to article comments. This means that an endpoint that returns "all comments made by a user" actually returns "all comments made on Help Center articles by a user."
If you have multiple brands, you must use the brand-specific subdomain in the API requests. For example, if you have a brand named brand1, a request for information about that brand's help center content might look like
The following is specific to the Zendesk Help Center API:
- Zendesk recommends using the string type for ids in Help Center.
- Locales are returned and passed as strings. Example: 'en-us'.
For more information on data types used in the APIs, see Data types in the general API introduction.
The Help Center API supports two methods of paginating through results: cursor pagination and offset pagination.
Offset pagination works as described in Using offset pagination in the general API introduction. Most Help Center list requests return 30 records per page by default when using offset pagination. You can change the number on a per-request basis by including a
per_page parameter in the request's query string. Example:
per_page=100. However, you can't exceed 100 records per page for most requests.
Cursor-based pagination is available in a subset of endpoints. For more information, see Using cursor-based pagination in the general API introduction.
The Help Center API follows the Support API's requests-per-minute limits. See Rate limits in the general API introduction. Requests in one API don't count against the rate limit of the other API.