Nano Banana Pro (Gemini 3 Pro Image)
Nano Banana Pro, built on Gemini 3 Pro, renders legible multilingual text, outputs up to 4K, blends up to 14 images and keeps 5 people consistent.
Reasoning-based image model with Search grounding for real-world facts such as recipes, weather and diagrams, plus studio-style controls. Ships with SynthID watermarks across the Gemini app, Ads, Workspace and AI Studio. GA 2026-05-28.
- Date
- Thursday, 20 November 2025
- Lab
- Google DeepMind
- Kind
- model
- Access
- closed API
Figures
| Measure | Value | Measured by |
|---|---|---|
| Max resolution | 4K up to 14 input images and 5 consistent people | company |
Also catalogued in the media file under the same id.
Sources
- deepmind.google/blog/introducing-nano-banana-pro/
- ai.google.dev/gemini-api/docs/changelog
- blog.google/technology/ai/nano-banana-pro/
- techcrunch.com/tag/image-generation/
This record was checked against its sources on 6 October 2026. How we check