Going forward, Google is introducing new features to be showcased at Google I/O. The new imaging model, Imagen 3, will be available now on Gemini, Gemini Advanced, Business, and Enterprise.
Example images generated using prompts – results may vary.
– Animated image of a baby dragon hatching from an egg, in a sunlit meadow, surrounded by exotic glowing butterflies. Vibrant colors, detailed scales.
– A realistic image of a stunning mountain landscape with jagged peaks and snow-capped peaks, bathed in the warm glow of sunset. Dramatic clouds, painted in vibrant shades of orange, pink and purple, stream across the sky, casting long shadows across the rugged landscape.
– A photo of a fluffy dress made of tissue paper in an elegant showroom.
– Vibrant abstract painting with the phrase “Dream Big” spread across the canvas in bold colors.
Imagen 3 sets a new standard in image quality, creating images with just a few words. You can also ask Gemini to create images in a variety of styles – such as realistic landscapes, ornate oil paintings, or whimsical clay.
Caption: Image 2 vs. Image 3
Imagen 3 offers advanced imaging capabilities that come with built-in protections and adhere to product design principles. Across a wide range of criteria, Imagen 3 performs favorably against other available imaging models. As with Imagen 2, we use SynthID, Google’s digital watermarking tool for AI-generated images.
The design principles are clear: from start to finish, the user remains in control of the creative process. If the initial image obtained does not meet expectations, simply ask Gemini what you would like to change and a new image will be created.
Over the next few days, Google’s Gemini platform will also make it possible to create images of people, with early access being released to Gemini Advanced, Business, and Enterprise users, starting in English. We’ve been working on technical improvements to the product, as well as improved evaluation groups, red team exercises, and clearer product principles.
Google says it doesn't support creating realistic images of identifiable individuals, depicting minors, or excessively gory, violent, or sexual scenes.
“Of course, not every image Gemini creates will be perfect, but we will continue to listen to feedback from users who have early access to Gemini Advanced as we continue to improve. We will roll it out gradually, with the goal of rolling it out to a wider range of users and languages soon,” the American multinational company says.
To be among the first to try these new features, try them today. Advanced Gemini Or sign up for Gemini for Workspace.
“Coffee trailblazer. Social media ninja. Unapologetic web guru. Friendly music fan. Alcohol fanatic.”