Nano Banana
Use Google’s Nano Banana models for product shots, consistent characters and AI edits. Learn references, prompting and resolution.
Last verified
Best for
Use Google’s Nano Banana models to place products and characters in new scenes or edit images with a prompt. Nano Banana 2.1 is the default for new accounts and accepts up to 14 references. It replaced Nano Banana 2 on October 6, 2026; the examples here used that earlier model. Pro offers finer detail and text.
For long exact copy and dense layouts, try GPT Image. For posters and logos, try Ideogram. For files you can scale and edit, use Arrow. Which model for what shows all five families on the same prompts.
Which version
| Model | Best for | Credits per image | Reference images |
|---|---|---|---|
| Nano Banana 2.1 | Pro quality at Flash speed, up to 4K | 10 | Up to 14 |
| Nano Banana 2 Lite | Google's fastest and cheapest, built for iterating | 8 | Up to 14 |
| Nano Banana Pro | Highest quality from Google, up to 4K | 35 | Up to 14 |
Start on Nano Banana 2.1. Explore on Lite, at 8 credits and about 5 seconds an image, then run the prompts you like again on Nano Banana 2.1. Switch to Pro for its finer detail and text, as in an infographic or packaging. The original Nano Banana left Batch Banana on October 6, 2026; a request that names it runs on Nano Banana 2.1.
The same prompt on each version:
Write the prompt
Describe the scene in sentences. Include the subject, action, setting, composition and style. In this example, keywords produced a hare in the center; a full sentence placed it in the lower left with the requested olive coat.
Explain what the image is for. Name the brand, product and visual style. Our vague logo prompt invented a business and tagline. A specific brief followed the requested design, though it still added extra text.
Quote exact text and describe the lettering. Supply the final wording, font style, color and placement. Add "No other text" if extra copy would be a problem.
Give each reference a role. Identify each image and say how to combine them. Use consistent names for recurring characters or objects:
Fern is the brown hare in the olive coat. Image 2 is the Halvard chocolate bar. Put Fern on a ferry bench, breaking a square off the bar, in the same ink and watercolour style as Fern's picture.
More tips
Name the change and what to preserve. Specify the target color or detail, then list anything that should stay the same. "Make the wrapper green" let the model choose the shade; a more precise prompt produced pine green.
Describe framing and light. Terms such as "low-angle shot", "shallow depth of field" and "macro lens" give more direction than "masterpiece" or "ultra detailed".
Rewrite the prompt, don't reply to it. Each Create starts fresh, so the model doesn't remember your last prompt. When a result is close, copy the prompt, change the part that was wrong, and run it again.
Settings
Resolution appears on Nano Banana 2.1 and Pro. Lite has no size setting. Every version offers all ten aspect ratios, from 21:9 to 9:16.
| Model | Credits per image | Other settings |
|---|---|---|
| Nano Banana 2.1 | 10 | Resolution: 2K 15, 4K 23 |
| Nano Banana 2 Lite | 8 | Same price |
| Nano Banana Pro | 35 | Resolution: 4K 70 |
On Pro, 1K and 2K both cost 35 credits, so make it at 2K. On Nano Banana 2.1, draft at 1K for 10 credits and make the final at 2K for 15 or 4K for 23.
The aspect ratio you pick holds even when your reference images are a different shape. A 16:9 request with a square reference came back 16:9.
Limits
- Text can come out wrong. In our comparison, Nano Banana 2 printed copy of its own on a wrapper, and one small line of it is not a word. Check every word before you use an image.
- Every version reads up to 14 references.
Sources
Google's own guides: the Nano Banana prompting guide on Google Cloud and DeepMind's Gemini image prompt guide.
Images made in Batch Banana on October 2, 2026. Select Prompt on an image to read the prompt and model that made it. Where a run made more than one image, we show the one we liked best.