Google Drops Nano Banana 2.1: Cheaper, Faster, and Only Slightly Hallucinating
Google just released Nano Banana 2.1, and it is supposedly the budget-friendly genius we all needed. While it claims to outshine its predecessor with fancy mask editing and character consistency, we all know it is just another shiny toy from the Gemini factory.
The new Nano Banana 2.1, built on Gemini 3.6 Flash, replaces the aging Nano Banana 2. Google is betting hard on the idea that users want cheaper 1K image generation, cutting the price for Gemini API calls in half. The model now handles up to 14 reference images to keep characters and objects consistent across different scenes, which is a massive upgrade for anyone tired of AI-generated people having their faces melt between frames.
Technical specs allow for massive 8:1 aspect ratios and three tiers of "thinking" modes—minimal, medium, and high—integrated directly into Google AI Studio and Gemini Enterprise. It officially claims 5th place in the Arena leaderboards, narrowly beating its predecessor by a few dozen points in complex editing tasks.
However, Google admits that at 1K resolution, the text often turns into a blurry mess and instructions for masking are still more like suggestions than rules. Nano Banana Pro remains the heavyweight champion for complex tasks, as the 2.1 version tends to struggle with basic object scaling that a toddler could handle.
It is truly fascinating to watch the tech giants race to the bottom on price while trying to mask the fact that these models still can't draw a hand without requiring an exorcism. If the goal is to flood the internet with slightly higher-quality hallucinations for pennies, this is undoubtedly a monumental success.
Source: GPTunneL
Comments
Help shape the next version: Add context or suggest a correction. AI review can add points toward a rewrite. Reviews and updates may take time; a full meter does not guarantee a new version.