Users of the free version of Google Bard, Google's AI chatbot competitor to ChatGPT, can now take advantage of another ChatGPT-like feature: AI image generation.

Google added the ability to generate images to Google Bard last week, making it incredibly easy to jump in and tell the chatbot exactly what you want to see. The system runs on Gemini Pro and Google's Imagen 2 model to generate the images, meaning you can expect a decent balance between quality and speed. Google says that Bard should deliver "photorealistic outputs" when you use it, though the quality has definitely varied in my experience.

Creating images in Google Bard is like telling it to create an image for you. All you have to do is go to the Bard website and type "create an image" or "generate an image" and follow that with the image you want to see.

The results of these prompts can vary, as with any AI generation system, and it will sometimes take some configuring to find the perfect prompt for what you want the chatbot to produce. But with a little patience, you can usually get a decent result. Google’s Imagen 2 system is pretty impressive, even compared to some of the other image generation models we’ve seen, like Midjourney . What makes image generation in Google Bard particularly nice is that it typically spits out a few different style options for each prompt, which can cut down on the back and forth.