Hey Folks!
I am working on project where a client requested images be generated with Nano Banana via Gemini. The reasoning, the images seem to have more detail.
As I started working with the project, I noticed Gemni having issues with prompts that "worked" in ChatGPT:
- Using 3 character reference sheets and a background image.
- Repeating image reference prompts failed
Note: I was working inside a NotebookLLM.
I switched to a new chat outside of my Notebook, and still experienced issues.
Maybe I need to spend more time learning how to prompt in Gemini, because no LLM is the same.