AI - Google releases Nano Banana 2.1
Nano Banana 2.1 is here, and this looks like an update to the existing Nano Banana 2 rather than a completely new model. Google is pitching it as a faster, more efficient image workhorse, but with better quality and some useful improvements to editing and text rendering. Though is it really a significant upgrade given that it is still based on Gemini 3.6 Flash? 🤣
Nano Banana 2.1
Google AI Studio on X
Official announcement
Nano Banana 2.1 - Model Card — Google DeepMind
Nano Banana 2.1 is based on Gemini 3.6 Flash.
For more information about the model architecture for Gemini 3.6 Flash, see the Gemini 3.6 Flash model card.
What is new?
According to Google’s Gemini API documentation, the main improvements include:
- Better visual quality and realism at 1K, 2K and 4K resolutions.
- Fewer tiling artifacts in very wide and panoramic images.
- Improved text rendering and infographic layouts.
- The ability to combine up to 14 reference images, with consistency for up to four characters and fidelity for up to ten objects.
- Grounding with Google Web and Image Search.
- Configurable thinking levels.
The reference image limit is a pretty big jump, and better text in images and infographics is always welcome. The wide aspect-ratio fixes are a more specific improvement, but potentially a very useful one if you are generating banners or panoramas.
This is still positioned as the efficient option, with Nano Banana Pro remaining Google’s higher-end model for more complex image tasks. The model is available in the Gemini API as gemini-nano-banana-2.1, and Google recommends it for new projects.
I have not had a chance to try it yet, so we will have to see how much of an improvement it is in practice 😊
Relevant X posts
Useful links
#Blog #AI #Google #Nano Banana 2.1 #Google Deepmind #Nano Banana #Gemini Nano Banana 2.1