Nano Banana 2.1: better prompt accuracy, text and consistent characters
Google’s new image generation and editing model is generally available in the Gemini API. The previous version is already deprecated.
On October 6, Google made Nano Banana 2.1 generally available in the Gemini API. It updates the Nano Banana 2 model for generating and editing images.
What’s new
According to the Gemini API changelog, the model improves:
- overall visual quality;
- prompt adherence;
- character consistency across multi-turn conversations;
- text rendering;
- wide and panoramic aspect ratios at 1K, 2K and 4K.
How to try it
The API model ID is gemini-nano-banana-2.1. If you use the previous gemini-3.1-flash-image, plan your migration: it’s already deprecated, although Google hasn’t announced a shutdown date.
Limitations
The changelog has no specific quality numbers or competitor comparisons, so test the improvements on your own prompts, especially text in non-Latin scripts, which Google doesn’t address separately.
Prepared with AI from public sources.