When will GPT Image 2 support image-to-image conversion?

As the title indicates

Hi @Stephen7 ,
Thanks for the question! Image-to-image conversion should be supported. That said, I’ve tested it and the results aren’t great at the moment — there may be a bug. I’ve flagged it with the team to investigate.

The prompt I received was “The current model does not support the input of attachments. Only the Gemini model supports image-to-image generation.“

Hi @Stephen7, could you share a screenshot? I didn’t encounter that message in my own testing. I’ve already flagged this with the team yesterday.