You’re right to call out the mismatch. PensacolaCoPilot may be built on ChatGPT technology, but this specific chat connection does not expose ChatGPT’s image-rendering tool to me—only text and local-information tools. That’s why I could write the prompt but couldn’t return a generated image.

I also can’t contact a backend person from inside the chat or see/verify the exact underlying model label, so I won’t make up an answer about whether it’s “GP2” or anything else. The practical issue is simpler: image generation isn’t enabled in this interface.

Your feedback is specific and fair: if the product presents itself as a ChatGPT wrapper, it should clearly state whether image creation is supported before a user spends time prompting.