It's possible to create and manage your content using the Ghost Admin API. Our content management interface, Ghost Admin, uses the Admin API - which means that everything Ghost Admin can do is also possible with the API, and a whole lot more!

Secure authentication is available either as a User with role-based permissions, or as an integration with a single standard set of permissions designed to support common publishing workflows.

The API is RESTful with predictable resource URLs, standard HTTP verbs, response codes and authentication used throughout. Requests and responses are JSON-encoded with consistent patterns and inline relations and responses are customisable using powerful query parameters.

API Clients

JavaScript Client Library

We've developed an API client for JavaScript, that simplifies authenticating with the Admin API, and makes reading and writing data a breeze. The client is designed for use with integrations, supporting token authentication and the endpoints available to integrations.


Base URL


All Admin API requests start with this base URL.

Admin Domain

Your admin domain can be different to your main domain, and may include a subdirectory. Using the correct domain and protocol are critical to getting consistent behaviour, particularly when dealing with CORS in the browser. All Ghost(Pro) blogs have a * domain as their admin domain and require https.


Version strings are required and usually start with v. The api versioning guide explains the current available versions and stability index. The Admin API also has a stability index for specific endpoints.

JSON Format

The API uses a consistent JSON structure for all requests and responses:

    "resource_type": [{
    "meta": {}
  • resource_type: will always match the resource name in the URL. All resources are returned wrapped in an array, with the exception of /site/ and /settings/.
  • meta: contains pagination information for browse requests.

Composing requests

When composing JSON payloads to send to the API as POST or PUT requests, you must always use this same format, unless the documentation for an endpoint says otherwise.

Requests with JSON payloads require the Content-Type: application/json header. Most request libraries have JSON-specific handling that will do this for you.


All browse endpoints are paginated, returning 15 records by default. You can use the page and limit parameters to move through the pages of records. The response object contains a meta.pagination key with information on the current location within the records:



Query parameters provide fine-grained control over responses. All endpoints accept include and fields. Browse endpoints additionally accept filter, limit, page and order. Some endpoints have their own specific parameters.

The values provided as query parameters MUST be url encoded when used directly. The client libraries will handle this for you.

NOTE: This content is copied from just for demo purpose