Google is set to launch Nano Banana 2, a marked improvement over its predecessor, Nano Banana 1, following positive feedback from users of the Gemini app. The new model enhances capabilities by efficiently processing complex tasks like precise coloring and correcting textual elements in images. Recent leaks suggest that Nano Banana 2 features a multi-step workflow for high accuracy, including image planning, generation, review, and iterative correction. This model supports various aspect ratios and resolutions up to 4K, making it suitable for both creative and professional applications. Although it currently uses Gemini 2.5 Flash, future upgrades to Gemini 3.0 Pro are expected. A rebranding to Nano Banana Pro is also in consideration, aiming for higher accuracy and consistency. With internal testing underway and the imminent announcement, Google is expanding its AI toolset, offering users broader creative controls and enhancing workflows across platforms, including upcoming integrations in new experiments.
Source link
