Google employees have confirmed to outside developers that the experimental generative-image model known as “Nano-Banana” is a company project, according to posts shared late Tuesday. The acknowledgment marks the first public link between the internet giant and the model, which has been circulating on image-generation forums under an anonymous label. Nano-Banana has attracted attention for delivering rapid, high-quality edits and handling complex text prompts while using relatively modest computing power. Developers who have tested the system say its performance rivals larger, cloud-based models despite its smaller size. Testers report that early builds of Google’s forthcoming Pixel 10 smartphone include image-editing features powered by Nano-Banana, indicating the company is preparing on-device deployment. While Google has not formally announced a launch date or detailed specifications, the confirmation underscores the firm’s push—led by Google DeepMind—to miniaturize generative AI for consumer hardware.