Google’s Nano Banana 2.1 promises better images for about half the cost

Google has released Nano Banana 2.1, an image generation and editing model that replaces Nano Banana 2 and cuts prices by roughly half. Matthias Bastian reports for The Decoder that the new model is based on Gemini 3.6 Flash and is rolling out across Google’s consumer and business AI products.

Google says Nano Banana 2.1 improves image quality, text rendering, infographic creation, editing, and consistency when users work with recurring characters or products. The model can use up to 14 reference images in a request and is designed to maintain up to four characters and ten objects across generated images.

The lower pricing is one of the most notable changes. A 1K image costs 3.36 cents, compared with 6.70 cents for Nano Banana 2. A 4K image costs 7.56 cents, down from 15.10 cents. Google’s higher-end Nano Banana Pro still costs 13.40 cents for a 1K output.

New controls and wider availability

Nano Banana 2.1 supports 1K, 2K, and 4K image generation at Flash-tier speeds. Users can select minimal, medium, or high thinking levels, which Google says can affect output quality. The model also connects to Google Search, potentially allowing it to draw on current web information for relevant requests.

Google’s benchmark results place the model ahead of Gemini 3.1 Flash Image and Gemini 3 Pro Image in areas including text-to-image generation, infographics, general editing, and multi-reference editing. However, benchmark scores do not necessarily settle practical quality differences. The Decoder notes that Nano Banana Pro often continues to create more realistic and natural-looking images, particularly in color and proportions.

Nano Banana 2.1 is becoming available in the Gemini app, Google Search AI Mode, Google AI Studio, Flow, Stitch, Google Ads, and the Gemini Enterprise Platform. Google plans to shut down the previous gemini-3.1-flash-image model on October 29, 2026.