How to prompt Nano Banana 2

How to structure your prompt

A natural-language request that starts with Create, Generate or Draw, names the image type and style (for example photorealistic, watercolor painting, kawaii sticker), then describes the subject, what it is doing and the setting in full descriptive sentences. For photos add the lighting, shot type, camera angle and lens. For text, give the exact words in quotation marks and describe the font style in words; add 'No other text' when only that text should appear. State the aspect ratio in ordinary words at the end.

Length and syntax

Up to 2,000 characters. Longer prompts are not better if the extra words do not change what appears on screen.

Exact in-image text in quotation marks with the font described in words (for example 'in a clean, bold, sans-serif font'). Aspect ratio in plain words ('Aspect ratio 16:9', 'Square image'). No parameter flags.

Prompt skeleton

Create a [image type and style] of [subject with key details] [action] in [setting]. [Lighting]. [Shot type, camera angle and lens, for photos]. [The text "..." in a [font description], placed [position]. No other text.] [Aspect ratio in words].

What it’s best at

• legible text in images, including non-English languages
• infographics and diagrams using real-world knowledge
• photorealistic product and commercial shots
• stylized illustrations and stickers
• character consistency across images

Known limitations

• text and fine details still need human review
• in the Gemini app, Nano Banana 2 is used only when the Gemini model is set to Flash or Pro; Flash-Lite uses Nano Banana 2 Lite
• editing an existing image is a separate workflow

Avoid in your prompt

• Midjourney parameter flags
• unrequested text in the image
• vague single-phrase prompts when specific details matter