Google has announced the integration of its latest image generation AI model, Imagen 3, into the Gemini platform, making it available to all users, including those on the free version, and supporting all languages, including Japanese. This new feature, which requires English prompts for now, will soon resume the generation of human images for paid Gemini users (Advanced, Business, Enterprise). The model prioritizes safety by not supporting the creation of realistic, identifiable individuals, minors, or excessively violent or sexual scenes. Additionally, Google introduced a new chatbot creation feature called "Gem" for paid Gemini users. Imagen 3, unveiled at this year's Google I/O, is already available in the U.S. through the "Pixel Studio" app on the Pixel 9 series. The model boasts significant improvements in visual richness, texture accuracy, and noise reduction, enabling the generation of diverse styles such as realistic landscapes, detailed oil paintings, and playful clay animations. It also excels in understanding complex prompts and accurately rendering text in images, such as birthday cards and presentation materials. To ensure safety, strict content filtering and the "SynthID" watermark for identifying AI-generated images are implemented. These advancements are expected to enhance user experience significantly, although not all generated images will be perfect.