Google yesterday launched Nano Banana 2, the consumer brand for Gemini 3.1 Flash Image. The upgrade includes sub-second 4K image synthesis across multiple aspect ratios, character consistency for up to five characters, fidelity for up to 14 objects in a single workflow, and precise text rendering accurate enough for marketing mockups and greeting cards. The images all carry SynthID watermarks.
Nano Banana 2 is available through the Gemini API, Gemini CLI, Vertex API, AI Studio, and Antigravity, and it’s also the default image generator in Search results via Google Lens and AI Mode across 141 countries. This makes it ambient, like spell check or autofill for billions of users.
A year ago, AI image generators couldn’t render legible text on a birthday card. Today, they’re producing print-ready marketing assets with brand-consistent characters at sub-second speeds. Stock photography libraries, product shot studios, and social media content farms all sell labor that Nano Banana 2 gives away for free. Some will call it AI Slop, others will call it AI Candy, and still others will recognize it for what it is: the next generation of imaging tools.
I did not mourn the loss of my Rapidograph pen, my non-repo blue pencil, my single-edge razor blades, my red-coat “cake opaque,” or my roll of Rubylith. If you don’t know what these things are, it’s because the personal computer ended the need for them (for paste-up and pre-press) decades ago. Every single person who wanted to keep working in an art department learned to do the job with a new, digital co-worker (mostly Mac, but occasionally a PC).
Author’s note: This is not a sponsored post. I am the author of this article and it expresses my own opinions. I am not, nor is my company, receiving compensation for it. This work was created with the assistance of various generative AI models.