Skip to main content
Google Cloud debuts its latest generative AI model.

Unveiled today alongside the API for Google’s Gemini AI model, Google says its Imagen 2 text-to-image generator delivers “significantly improved image quality” compared to its predecessor. This includes rendering accurate text (instead of a jumbled mess) and logos that can be overlayed onto products and surfaces.

That should bring it more in line with AI products from rivals like Adobe’s Firefly and Bing’s DALL-E integration, which have also seen similar updates in recent months.

Images generated using Google Cloud’s new Imagen 2 AI model.


If you specify what text you want to see generated within your prompts, Imagen 2 should follow those instructions to the literal letter.
Image: Google