Google releases Imagen 3 for image generation to all Gemini users: Details

Google said that Imagen 3 is its highest quality image generation model, bringing an even higher degree of photorealism, better instruction following, and fewer distracting artefacts

Image generation with Imagen 3 is now available for all Gemini users. Imagen 3 was previewed at I/O 2024 in May and announced in August along with Gems. Google has stated that Imagen 3 offers greater versatility, improved prompt understanding, higher quality images, and better text rendering.

“Imagen 3 is our highest quality image generation model yet, bringing an even higher degree of photorealism, better instruction following, and fewer distracting artefacts than ever before,” said Google in a post on X by the Google Gemini App.

To access the features, users must provide prompts that start with "draw," "generate," or "create," while mentioning the desired style, such as photorealistic, watercolor, painting, or cartoon illustration. Gemini informs users that it is creating an image with Imagen 3, and users can download the image with a SynthID watermark in place. Google describes SynthID as “SynthID watermarks and identifies AI-generated content by embedding digital watermarks directly into AI-generated images, audio, text, or video.”

As of now, the generation of images of people is not available for free users. Google began making this feature accessible to Gemini Advanced, Business, and Enterprise users in English during early access in May.

In related news, Google is launching the "Ask Photos" feature for select users in the US. Powered by Gemini, this feature is being rolled out as a server-side update within the Google Photos app, meaning it is not dependent on the version of Google Photos users have. Ask Photos allows users to ask questions in natural language directly within Google Photos, leveraging Google’s Gemini AI models.