GET /v1/resources

Summary

Returns the set of available models and their capabilities. Use this endpoint to discover which models are available, what modality they support, and which run options they accept.

The response includes a content hash that can be used for cache invalidation. CLI tools send the hash via the cli-agent header; if the hash no longer matches, the server sets an action-require response header signalling that the client should refresh its local cache.

HTTP Request

GET /v1/resources

Response

Success (200 OK)

Returns a JSON object whose status is "SUCCESS" and a data object containing:

Field	Type	Description
hash	string	Base64url-encoded content hash. Use this to detect changes in available models.
resources	object	Container for available resources.
resources.models	array of objects	List of available models.
resources.models[].model	string	Model identifier (e.g. "gpt-4o", "dall-e-3").
resources.models[].modality	string	Model modality: "text", "image", or "video".
resources.models[].availableRunOptions	array of objects	Run options this model supports. Empty array if none.
resources.models[].availableRunOptions[].key	string	Option key (e.g. "reasoningEffort", "quality").
resources.models[].availableRunOptions[].options	array of strings	Accepted values for this option.

Example response

{
  "status": "SUCCESS",
  "data": {
    "hash": "dGhpcyBpcyBhIGhhc2g",
    "resources": {
      "models": [
        {
          "model": "o3",
          "modality": "text",
          "availableRunOptions": [
            {
              "key": "reasoningEffort",
              "options": ["low", "medium", "high"]
            }
          ]
        },
        {
          "model": "gpt-image-1",
          "modality": "image",
          "availableRunOptions": [
            {
              "key": "quality",
              "options": ["low", "medium", "high"]
            }
          ]
        },
        {
          "model": "gpt-4o",
          "modality": "text",
          "availableRunOptions": []
        }
      ]
    }
  }
}

Error responses

HTTP status	status code	Description
401	UNAUTHORIZED_ERROR	Missing or invalid API key.

See also

• Models & Options guide: Models & Options
• Run options reference: Run options