For the GPT image 1.5 model, we are currently unable to upload images to open ai got image models though api for editing. Feature Request: Add image edit endpoints and support for image uploads to these models.
Currently, AI Gateway supports image generation through multimodal models (like GPT-5 variants) and dedicated image models (like FLUX and Imagen), but doesn’t yet have endpoints for uploading and editing existing images.
This would be a great addition! The team have been told 