Nano Banana 2.1: cheaper 1K images than Nano Banana 2, and no API free tier

Google published the Nano Banana 2.1 model card on October 6. It is an image generation and editing model based on Gemini 3.6 Flash, and the Gemini API page lists it as gemini-nano-banana-2.1, an update to Nano Banana 2.

Where it runs

The model card lists the Gemini app, Google AI Studio, the Gemini API, AI Mode in Search, Google Ads, Google Flow, and Google Stitch.

What changed

From the API page:

• 1K, 2K, and 4K output, 1K by default
• Fixed tiling artifacts on very wide and tall aspect ratios at 2K and 4K
• Up to 14 reference images, with character consistency for up to 4 characters
• Grounding with Google Web and Image Search
• Thinking levels: minimal, medium (default), high
• Batch API supported; caching and function calling not supported

In Google's side-by-side human preference tests, Nano Banana 2.1 with thinking scored 1050 for overall text-to-image preference, against 990 for Nano Banana 2 and 935 for Nano Banana Pro. Google lists known limits including blurry small text at 1K, imperfect character consistency, and left/right confusion. It also says the model is "still limited" in 3D reasoning.

What it costs

The Next Web reports, from Google's price list, $0.0336 per 1K image against $0.067 for Nano Banana 2, and $0.113 per 4K image against $0.151. It reports input at $1.50 per million tokens, up from $0.50, and no free tier. We did not check Google's pricing page directly for this draft.

Should you switch

If your pipeline outputs 1K or 4K images from Nano Banana 2, the reported per-image price is lower. If you rely on the free tier or send large text inputs, the reported changes go the other way.