- Published on
Google Launches Nano Banana 2.1: Better Than Its Predecessor at Half the Price
- Authors
- Name
- DeepHorus
Google on Oct. 6 released Nano Banana 2.1, the newest version of its AI model that turns a typed sentence into a picture or edits an existing photo. It is rolling out in the Gemini app, Google Search's AI Mode, Google Ads, and developer tools including Google AI Studio, Flow and Stitch. Google says the upgrade outperforms its previous models across the board.
The original Nano Banana made Gemini a hit. In September 2025, its trick of turning selfies into collectible figurines pushed Gemini to No. 1 on both major app stores, dethroning ChatGPT after nearly three years, while Alphabet's market value topped $3 trillion. Version 2 arrived in February, built on Gemini 3.1 Flash Image, and could check Google Search before drawing real-world subjects.
Google points to three upgrades in 2.1: better visual design, sharper mask-based editing where only a marked region changes, and stronger subject consistency across repeated edits.
On overall preference in text-to-image tests, Nano Banana 2.1 scored 1,050 ELO points, versus 990 for Nano Banana 2 and 935 for Nano Banana Pro. Google's tests also gave it an infographic factuality score of 0.521, against 0.179 for its predecessor, though independent testing was not available at launch.
Technically, it handles up to 14 reference images, tracking four characters and ten objects, and outputs up to 4K in formats as wide as 8:1. Developers can set how long it "thinks" before drawing and let it ground results in Google Search and Google Image Search.
Pricing is roughly half of Nano Banana 2's: a standard 1K image costs $0.0336 through the API, down from $0.067, while a 4K image costs $0.0756 versus $0.151. Batch jobs receive another 50% off, putting 1,000 images at about $33.60 instead of $67.
The cheaper, faster model could push AI image generation deeper into advertising and third-party apps, though Google's own benchmarks will need independent confirmation.
