How we struggled with generating text in images. AI drew scribbles instead of letters, and captions for postcards had to be added by hand?

Yesterday, Nano Banana 2 (official name — Gemini 3.1 Flash Image) was announced - an updated model - now image generation is not just fast, but truly useful for work.


What's new?
First, text in images is now perfectly readable. The model can not only write but also translate text directly inside the image - convenient for international projects.

Second, infographics and diagrams: you can turn boring notes into clear data visualizations.

Third, control and consistency - you can save up to 5 characters and 14 objects so they look the same across a series of images (goodbye, inconsistency in storyboards!).

And also up to 4K resolution, support for complex prompts, and integration everywhere: in Google Search, Lens, AI Studio, and even Google Ads.
📌 By the way, in Flow (a platform for creators), Nano Banana 2 became the default model with zero credits.

So you can test it for free)

👉 Grab it here