Skip to main content

​
Introduction

GPT-Image is OpenAI’s image model family for text-to-image generation, editing, restyling, and blending reference images. These documents currently cover image-2, image-2.6 Flare / Sunburst, and image-2.5 Flare / Sunburst, with separate guides for each endpoint.
  • Base URL: https://api.gravitex.ai
  • Authentication: Authorization: Bearer sk-xxxxxxxxxx
  • Content-Type: application/json for generation; JSON or multipart/form-data for editing

​
Documented models

​
Supported endpoints

The image-2 guide documents gateway extensions for image input on generations and text-only edits. Do not assume those extensions apply to image-2.6: use generations for text-to-image and edits with images for editing.

​
Response and billing

Both endpoints return Base64 images in data[].b64_json and token statistics in usage. Text input, image input, and image output are billed separately. See endpoint guides for details.

Latest models and pricing

Check the console model catalog for current models, capabilities, and prices.

​
Integration flow

  1. Obtain an API key and configure the gateway URL and Bearer authentication.
  2. Choose the generation or editing endpoint guide.
  3. Send the selected model ID, prompt, and the desired parameters; include reference images for editing.
  4. Decode data[].b64_json to save the images and read usage for token consumption.