Firefly models don´t portray the picture that is described in the keywords

Question

I subscribed to Firefly Pro (approx. €22) a few months ago; back then, it was possible to use Firefly models to create images that closely matched the description (keywords) entered in the prompt—for example, an image of St. Stephen's Cathedral.
Now, however, when I enter keywords like "St. Stephen's Cathedral at St. Stephen's Square, view of the entire cathedral." The Firefly models display churches that do not actually correspond to St. Stephen's Cathedral.
The Gemini 3.1 (Nano Banana 2) model renders a more realistic depiction of St. Stephen's Cathedral.
Why is that so? That must be a mistake in the Firefly models. Please help me.

N8. · Accepted Answer

Hey ​@responsible_thinking125dAdobe Firefly generally has restrictions against producing images of many famous, trademarked, or copyright-protected landmarks (like the Eiffel Tower or Empire State Building) to maintain commercial safety. If it does produce a landmark, it is usually because the location is in the public domain, is a generic architectural style, or falls outside Adobe's restricted keyword filters.CheersNate

Sign up

To post, reply, or follow discussions, please sign in with your Adobe ID.

Sign in to Adobe Community

To post, reply, or follow discussions, please sign in with your Adobe ID.