Google introduces Imagen 3, its highest-quality text-to-image model, available in private preview

Join us in returning to NYC on June 5th to collaborate with executive leaders in exploring comprehensive methods for auditing AI models regarding bias, performance, and ethical compliance across diverse organizations. Find out how you can attend here.

Google is working on an update to its text-to-image model, saying the new Imagen 3 will provide an incredible level of detail, better understand natural language, and offer better text rendering. It’s available for select creators in private preview in ImageFX, but developers can sign up for a waitlist. Imagen 3 is said to be coming to Vertex AI as well.

“It’s our most capable image generation model yet,” Douglas Eck, senior research director of Google DeepMind, says on stage during the company’s I/O developer conference. “Imagen 3 is more photorealistic, with richer details and fewer visual artifacts or distorted images. It understands prompts written the way people write—the more creative and detailed you are, the better. And Imagen 3 remembers to incorporate small details…in longer prompts. Plus, this is our best model yet for rendering text, which has been a challenge for image generation models.”

WM Clay sculp.max
Example of Imagen 3. Image credit: Google

Today’s model announcement comes six months after Google made Imagen 2 generally available on Vertex AI. The company also recently rolled out text-to-live capabilities for the model in April. However, the work is not insignificant—it’s critical for Google to remain a step ahead of its AI competitors, many of which have image generation tools of their own, such as OpenAI’s DALL-E, Midjourney, Adobe’s Firefly, Meta’s AI, Microsoft’s Designer and more.

Example of Imagen 3. Image credit: Google
Example of Imagen 3. Image credit: Google

But not everything has been rainbows and sunshine for Google in this space. The company came under fire earlier this year when its AI was accused of being “too woke” after its AI created images of the U.S. Founding Fathers featuring a Black man and generated images of German soldiers from World War Two with a Black man and Asian woman. Google has apologized, stating it missed the mark. CEO Sundar Pichai reiterated that message during an interview with Bloomberg’s Emily Chang.

VB Event

The AI Impact Tour: The AI Audit

Join us as we return to NYC on June 5th to engage with top executive leaders, delving into strategies for auditing AI models to ensure fairness, optimal performance, and ethical compliance across diverse organizations. Secure your attendance for this exclusive invite-only event.

Request an invite

The Imagen 3 reveal joins the many other headlines coming out of Google I/O, including the launch of a new video generation model called Veo.

Source link

About The Author

Scroll to Top