GPT-4o Image
Generate images using OpenAI GPT-4o model (async)
- Native Multimodal - Image generation built into GPT-4o architecture, not external model calls
- Precise Text Rendering - Significantly better text generation than DALL·E 3
- Multi-Turn Consistency - Refine images through natural conversation while maintaining style coherence
- Image Understanding & Editing - Support uploading images as input for understanding and transformation
Authorizations
Please use your API Token for authentication
Body
Task parameters
Size (required) - Image aspect ratio, must be one of the supported formats
1:1, 3:2, 2:3 "1:1"
File URL list (optional) - File URL list, max 5 images
[""]Enable enhancement (optional) - Prompt enhancement, default false
false
Mask URL (optional) - Mask image URL for editing, black areas will be replaced/modified
""
Number of variants (optional) - Number of image variants, options 1, 2 or 4, default 1
1, 2, 4 1
Prompt (optional) - Describe the content you want GPT-4o to generate
"A beautiful sunset over the mountains"
Response
OK
Task ID
"01234567-89ab-cdef-0123-456789abcdef"