Nano Banana 2.1 is Google's update to Nano Banana 2 (gemini-nano-banana-2.1), a Gemini image model that can generate and edit images. This page describes the image generation capabilities of the model. Generation is synchronous: you submit a prompt and the result comes back in a single call, so there is nothing to poll.
What it's good at
Nano Banana 2.1 turns a written description into a finished image. In generation mode, it suits:
- Prompt-driven generation from a blank canvas
- Clean, detailed single images from a clear description
- Everyday image generation
Settings
- Image size: sets the output pixel budget, chosen from 1K, 2K, or 4K. Default: 1K.
- Aspect ratio: sets the aspect ratio of the output image, fitted to 1:1, 16:9, 9:16, 4:3, 3:4, or 21:9. Default: 1:1.
These are independent controls. The size decides how many pixels you get, the ratio decides what shape they make. Nano Banana 2.1 has no 512 size; the smallest size is 1K.
Nano Banana 2.1 does not allow controlling the seed, guidance scale, or inference steps. The prompt is your primary control over the generation.
Practical tips
- Write a specific, concrete prompt: name the subject, setting, lighting, and style you want.
- Reach for a larger image size only when you need the detail. Higher resolution outputs cost more and take longer.
- To make changes to an existing image rather than make a new one, switch to the editing version of this model.
- For a 512 draft, use Nano Banana 2, which keeps that size.
