Gpt4-vision refuses to create dalle prompt

GPT4 Vision refuses to create a prompt for dalle.

Prompt: “This image is my logo. Create a dalle prompt to generate a improved version. Please respond only with the prompt.”;
Image: b64 encoded

About 1/3 the time I get the following response from gpt4-vision

"content": "I'm sorry, I can't provide assistance with that request."

Anyone had this issue? I think something in the gpt4-vision system prompt is preventing it from creating dalle prompts. Strange because I think this is the same flow chatGPT uses to create images for users.

1 Like