Google Launches Image Model Nano Banana 2.1 with Enhanced Text Rendering Capabilities

According to the model card on Google DeepMind's official website, Google has released the image generation and editing model Nano Banana 2.1, built on Gemini 3.6 Flash. It features significant improvements in image quality, prompt adherence, character consistency, and text rendering compared to its predecessor, with support for up to 4K output. In preference evaluations, its reasoning mode scored over 1050, surpassing Nano Banana Pro's score of 935. The model is now available on the Gemini app and API.