谷歌又一张“王牌”发布！网友玩疯了

Core Insights - Google has launched the Nano Banana Pro model, which shows significant improvements over its predecessor, Gemini 3, particularly in visual information generation and editing capabilities [1][4][24] Model Improvements - Nano Banana Pro offers enhanced resolution capabilities, supporting 2K and 4K outputs, which is crucial for professional users [4] - The model allows for better text generation, multi-language support, and direct translation within images [4] - Improvements in photography quality include adjustable shooting angles, focus, and color, with aspect ratios ranging from 1:1 to 9:16 [4] Performance Testing - Initial tests showed that while Nano Banana Pro improved detail accuracy, it still had some flaws, such as incorrect text and extra limbs in generated images [5][6] - Further testing indicated that the model could accurately generate magazine covers and integrate multiple images, although the results were still somewhat basic and lacked interaction between characters and scenes [13][15] User Experience - Users have reported positive experiences with the model's ability to understand complex lighting and material properties, making it beneficial for small to medium-sized sellers [17] - The model has also demonstrated the ability to distinguish between different styles of Taekwondo uniforms, showcasing its advanced recognition capabilities [19] User Access and Subscription - Google offers various user options for accessing Nano Banana Pro, including limited free access for consumers and students, while subscription users receive higher quotas [22] - Developers and enterprises can utilize the model through Gemini API and Google AI Studio for creating rich user experiences [24] Market Impact - Following the launch of the new model, Google's stock price reached an all-time high, with a market capitalization of $3.65 trillion, briefly surpassing Microsoft [25] - Despite a recent decline, Google's stock has seen an increase of over 50% this year, indicating strong market performance [25]