Complete Guide to Using nanobanana2
nanobanana2 is Google’s text-to-image model based on Gemini 3.1 Flash Image, offering Pro-level quality, Flash-level speed, and a price cut by about half. This article is the complete guide to using nanobanana2, taking you from entry to advanced usage in one go.
1. Entry and Registration
- Gemini App: Go to https://gemini.google.com/, log in, and select Nano Banana 2 or Fast mode to generate images.
- Dedicated Entry: https://app.nanobanana2.xin/generate/image-tools/nano, focused on nanobanana2, one-click access.
2. Text-to-Image and Image-to-Image
- Text-to-Image: In the input box, describe in Chinese using “subject + scene + style + lighting/ratio”, e.g., “A woman in Hanfu admiring snow in a Jiangnan garden, holding an oil-paper umbrella, cinematic lighting, vertical 9:16”.
- Image-to-Image: Upload a reference image and input modification instructions (e.g., “Keep the character consistent, only change the background to a snowy scene”). nanobanana2 will reconstruct while preserving the structure.
3. Prompts and Advanced Techniques
- Structure: Subject + Action/State + Environment + Style + Perspective + Lighting + Text Content (optional) + Resolution/Ratio.
- Subject Consistency: Can maintain consistency for up to 5 characters and 14 objects in a single session, suitable for serials and series.
- Multi-Style: Performs well with single styles like Pop Art, Japanese manga, LOMO, Chinese ink wash; can be combined with reference images for style uniformity.
4. Summary
Follow this complete guide to using nanobanana2 to go from zero to proficient. To experience nanobanana2 right away, click the button below.