I hope they take into account all the feedback I've sent and address the problems. to me the biggest issue with pro is the frequent face morphs from female to male while editing an image. flash, which was quite good until 3 pro came out, now is almost unusable (too many random images returned).

the issues I've been having with nano banana pro have not been solved, and now they reduced the daily allowance for pro users to 20 images (wtf!). nano banana flash is mostly throwing out random images with 0 relation with the prompt. maybe it's time to stop the sub and try something new.

I am so happy for you Ivan. so well deserved. you are at the very heart of our community my friend. congratulations and enjoy the best birthday ever indeed!

and also for the sake of clarity: the negative feedback is for image editing only. image generation from prompt is vastly improved and the 2400px resolution a very appreciated enhancement.

to be precise, the issues happen when you enable the reasoning model (Gemini 3 pro). if you keep using the fast model (2.5) images generated are still as good as before (but lower resolution of 1184px of course).

menu for today is first course on Langchain4j and second course on MCP servers with GitHub Copilot ☺️☺️