Imagen 3 is Google’s AI image generator, which was already announced in May at the company I/O Developer Conference. It launched in limited capacity in the US in August but became available for free Gemini users last month. I’ve been using it ever since to create all kinds of images, and while it’s an impressive tool overall, it has several limitations which hinders the overall experience.

This is where Imagen 3 struggles

The first limitation to be aware of is that you cannot generate images of people, at least with a free Gemini account. This applies not only to creating images of famous people, which many image generation tools don’t allow anyway, but people in general. So a prompt like “create a picture of two random people dancing” will return no results. For reference, ChatGPT also has this limit in place for its free tier.

However, you can create images of people – excluding famous people – if you choose a Gemini Advanced subscription. I tried it, and it’s a hit-and-miss. While it can generate images that are so realistic that it’s hard to tell if they’re AI-generated or not, sometimes the results it produces are subpar. Check out the two examples below. The one on the left appears very realistic and looks like it was taken by a professional photographer, while the other one just looks cartoony. Even when you asked the tool to make the photo more realistic several times, the changes were minimal.

Speaking of professional photographers, let’s move on to the other limitation or problem I see with the Imagen 3. Even when producing a realistic image, whether of a person, animal or object, the result looks professional rather than relaxed. Each shot is picture perfect, with the bokeh effect often added to make it look more appealing. Every image that Imagine 3 creates looks like it was heavily edited, which is fine if that’s the look you’re going for, but having the ability to make images look more casual would be nice.

I think sometimes the best photos are the ones that are raw. The unedited ones you took without much thought when the lighting wasn’t perfect and the people you captured didn’t even know you were taking a photo. This is where Imagine 3 struggles, although it’s worth mentioning that this is true of almost every AI image generator out there.

This brings me to the third major problem with Imagen, which is editing the created images. If I create a funny image of a cat wearing a hat and eating a popsicle and then want to edit it with an additional prompt, Imagen 3 will create a completely new image in Gemini. So, for example, if I like the image created but just want to change the color of the hat from black to blue, the tool will generate a new image entirely and change the color of the hat instead of just changing the color of the hat and leaving everything else as it is. Sure, the new image looks relatively similar to the old one when you use the right prompt, but it’s still not the same, which isn’t ideal. This makes it impossible to edit an image to perfection, especially with multiple prompts that will generate a new image each time. Check out the example below and see for yourself.

Another problem is that I can’t change the aspect ratio. Images are created in 1:1 aspect ratio by default and cannot be changed. If I tell the tool to change it to 16:9, Gemini just says it will but then generates a new image with the same aspect ratio. However, it looks like this will change soon, as the ability to change the aspect ratio is already in progress.

Limitations aside, Imagen 3 is good

Let me just make it clear that I’m not trying to knock Google’s fancy AI image generator. I just want to highlight the limitations I encountered while testing it so you know what to expect. Limitations aside, Imagen 3 is actually a very impressive tool. I’ve tried some of its competitors as well, and while every AI image generator has its pros and cons, I’d say Imagen 3 is among the best out there. My colleague Calvin agrees. He compared the tool with rivals and found that it is best available in terms of quality.

When Imagen 3 gets it just right, the results are outstanding. Pictures of animals, cities, people and anything else for that matter will be fantastic – if you can live with a photoshopped look. Don’t take my word for it. Take a look at the gallery below to see for yourself. And remember, we’re still in the early stages of AI-generated content, so just imagine what the software will be able to do in a few years.

Other limits to be aware of

These are the limits I ran into when I tested the tool and didn’t expect it – apart from the inability to generate images of people as a free user – although there are other limits in place that Google clearly states on its website. It’s worth listing them so you know what to expect.

Imagen 3 will not create an image it deems inappropriate, even with a paid plan. It includes images related to violence, harassment, sex, discrimination and the like. This also applies to images that encourage dangerous activity and those with harmful factual inaccuracies that would pose a risk to someone’s safety.

These are all appropriate limits, and most of the major AI image generation tools have them in place, not counting FLUX.1 used by Grok.