Hi Siry, Gaetan,
GPT-image-1 has been released in preview with limited access through the Azure OpenAI Service yesterday afternoon. This model brings major improvements over DALL·E, including better handling of detailed instructions, accurate text rendering in images, and the ability to edit images or perform inpainting by accepting images as input. It is now available to eligible users. To get started, users can follow the image generation how-to guide, and those interested can request access through the limited access model application.
For more information: GPT-image-1 released (preview, limited access)
I hope this information helps. Thank you!