Google has just rolled out a new update allowing all Gemini users to create images using its latest AI image model, Imagen 3. Whether you’re using Gemini on the web, in the mobile app, or integrated into Android, you can now access this powerful tool. Previously, image generation was only available to Gemini Advanced subscribers through the web version of the AI.
Unveiled earlier this year at Google’s I/O event, Imagen 3 was developed by Google’s DeepMind research lab. This AI model can produce highly realistic square format images, rivalling the outputs from other advanced platforms like Midjourney and Flux. The images are rich in detail and follow user instructions closely, making the tool incredibly versatile for creative projects.
However, there are some limitations for free users of Gemini. While everyone can use Imagen 3 to generate images, creating pictures of people is restricted. If you want to generate images that include people, you’ll need to subscribe to Gemini Advanced for $19 per month. Even with the advanced subscription, the AI won’t generate images of real individuals—focusing instead on fictional or generalized human representations.
One of the standout features of Gemini’s Imagen 3 is its ability to adapt and modify images based on simple text descriptions. This flexibility allows users to transform or completely change an image with minimal effort. If you don’t see the image creation feature yet, make sure to update the Gemini app to unlock these new abilities.
Imagen 3 is a significant upgrade from its predecessor, Imagen 2. The new model offers more realism and better detail, with smoother, more natural images that avoid the plastic-like appearance that users noticed in the older version. Google has improved the model’s capability to follow instructions, making it easier to generate the exact image you want.
Previously, Google had temporarily removed image generation from Gemini due to concerns about how the AI handled depictions of people. Now, with Imagen 3, the company is being more cautious while still delivering advanced image generation features. However, it still sticks to square images, unlike OpenAI’s DALL-E 3, which can create images in various orientations and sizes.
Though Imagen 3 is impressive, there are still some challenges. During tests, it occasionally refused to generate more complex images, such as someone sitting in a café on a moon base. Despite these limitations, for everyday image generation, Imagen 3 is an excellent tool for Gemini users looking to enhance their creativity.
Excited to try out Imagen 3 on Gemini? Share this article to let others know about the latest AI-powered image generation feature from Google!