As one of the most powerful text-to-image AI models, Google’s Imagen 3 has been available on Gemini apps, but only to an extent. Generating people was a luxury only afforded to paid users—until now.
You Can Now Generate Images of People Using Gemini’s Free App
If you use the Gemini web, desktop, or mobile app, you might have played around with its image generation capabilities, made possible through Google’s Imagen 3 model. Google released the tool to all users in August 2024, with options ranging from “photorealistic landscapes” to “whimsical claymation scenes,” according to the announcement in Google’s The Keyword.
However, generating images of people was another story. Google limited the roll-out of this specific feature to paid subscribers, including Gemini Advanced, Business, and Enterprise users. If you had a free account, requesting that Gemini produce an image including people garnered the following message:
“Generating images of people is only available in early access with Gemini Advanced. Get early access to new Gemini features when you subscribe to Advanced here.”
Now, as of February 2025, most free Gemini users can successfully generate images involving people on Gemini. That said, if you’re still getting an error message, the feature may still be on its way. It seems to be rolling out gradually, starting with US-based users who are using Gemini with English as the chosen language.
You still can’t produce realistic images of identifiable people or children, per Gemini guidelines.
How Does Imagen 3 Perform on Gemini Flash 2.0?
Also in February 2025, Google widely released its fastest Gemini model to free users. Google describes the model, called Gemini 2.0 Flash, as a “highly efficient workhorse,” which can only mean good things for image generation. Traditionally costing more energy and time compared to standard text generation, AI images can theoretically achieve a higher quality when a more powerful model as the engine.
Quality seems particularly important when it comes to depicting people. We all bore witness to early AI-generated humans, from warped hands shrouded in extra appendages to haunted, hollow eyes. So, you can understand my hesitation to trust Google’s AI-powered image generator as a free user.
Contrary to my fears, I was impressed by Gemini’s ability to handle my image generation prompts that included people. Particularly considering this service is free, I found the first few image attempts to be premium, realistic, and appropriate to the prompt provided.
You can ask Gemini to keep revising an image without having to start from scratch every time.
Being able to access Imagen 3 via Gemini for free has certainly been a fun tease, but having to avoid people was a significant limitation. Google’s latest update opens up the playing field, and makes Gemini’s free service a worthy competitor for paid services like Midjourney.
While paid services will have more advanced options, such as custom image sizes, Gemini’s approach is easy and ready to use. With so much available for free, I have to wonder if early access is still enough for people to pay for Gemini Advanced.
