What Nano Banana 2 does
Nano Banana 2 is Google's Gemini 3.1 Flash Image model. It creates images from a text prompt and edits existing ones from plain-language instructions — say 'add a sunset glow', 'replace the chair with a throne' or 'remove the background' and it applies the change. Attach reference images and it switches to editing automatically. Nano Banana 2 Lite is a lighter, cheaper variant.
- Intuitive Natural Language Editing
- Describe the change in plain language and Nano Banana 2 applies it — object swaps, background changes, facial refinements, style shifts.
- Text-to-Image Too
- Not only edits: describe a scene from scratch and Nano Banana 2 draws it; attach references and it edits instead.
- Precision Object & Background Editing
- Pixel-level object replacement, background modification, and facial refinement without distortions across iterations.
- Physics-Aware Outputs
- Edits maintain coherent lighting, natural shadows, and correct spatial relationships — outputs look real, not pasted.
How to use it
Upload Your Image
Upload the image you want to edit.
Type What to Change
Write what you want changed in plain language — 'swap the background to a forest', 'make the sky more dramatic'.
Get Your Edit
Receive your edited image in seconds. Iterate further with follow-up instructions.
Questions
Nano Banana 2 is Google's Gemini 3.1 Flash Image model. It generates images from text and edits images from plain-language instructions.
Nano Banana 2 (Gemini 3.1 Flash Image) is the faster, lighter model for everyday generation and edits. Nano Banana Pro (Gemini 3 Pro Image) is the heavier model for detailed output up to 4K with up to 8 references.
Object replacement, background changes, facial refinements, style shifts, lighting adjustments — all described in plain natural language.