I just tried Envato Elements’ new AI ImageGen — here’s what blew me away (and what didn’t)

I’ve been playing around with AI image generators for a while now, from Adobe Firefly to Stable Diffusion. So, when Envato Elements rolled out its latest AI ImageGen, I needed to test it.

They promise to blend the power of several generative AI models into one seamless experience—and they even throw in 5 free credits for anyone who signs up. I decided to take it for a spin. Could this be the all-in-one AI image tool creatives have been waiting for? Let’s find out together.

Hands on AI ImageGen

First of all, props to Envato for the name—I can’t be the only one who reads “ImageGen” as “imagine”. Right?

The interface is clean and straightforward: a prompt field and a big green “Generate” button. A settings icon lets you choose image orientation (square, landscape, portrait) and toggle the “Enhance prompt” feature, which uses AI to flesh out your ideas.

I started simple to gauge AI ImageGen’s capabilities. “Cool cat wearing a hoodie” seemed like a fun, simple test. In seconds, the tool spat out eight unique images. They weren’t groundbreaking—mostly cinematic shots with varying colors and styles—but they were all pretty close to what I envisioned.

This tool would prove useful if the goal were to see what different styles worked for specific themes. However, specific and targeted images would require significantly more input for prompts.

Next, I wanted to see how AI ImageGen handled human faces, notorious stumbling blocks for most AI art tools. I went with “hyperrealistic portrait of an old man, diversity, photograph”.

Before clicking “generate,” I flipped on the “Enhance Prompt” switch. The AI rewrote my prompt to “A hyperrealistic portrait of an elderly man with diverse ethnic features, capturing the depth of his wrinkles and wisdom in his eyes, set against a neutral background to emphasize his character.” Wow, it actually read my mind (or saved me at least two minutes of fine-tuning my prompt).

Out of the generated images, I only really liked the third option, rendered in monochrome. Surprisingly, despite specifying landscape orientation, the tool “smartly” created a vertical portrait within the horizontal canvas.

Now, for a real test of AI’s practical use: can it fool me into thinking a human food blogger styled a dish?

I typed in “delicious lasagna, foodie blog” and watched as the “Enhance Prompt” feature transformed it into “A mouth-watering lasagna with layers of rich, savory meat sauce, creamy béchamel, and perfectly melted cheese, styled for a foodie blog with warm, inviting lighting and rustic table setting.”

Pretty good description aside, the results were a mixed bag. While visually appealing with a cinematic flair, they had a distinctly AI-generated vibe—too polished, too perfect to mimic a genuine blog photo.

A casual scroller may believe a professional blog photo has been taken. For content creators, it may lack some personalization that would typically draw a larger appeal to their specific target audience and increase audience engagement through other unique choices that they make.

However, for stock photos, general purpose, or concept imagery, this quality is exceptionally useful due to a combination of its realism, stylized elements, and creative input (through writing longer prompts).

For my final test, I wanted to address a big question in the creative community: Can AI replace artists for specific artistic pursuits or at specific quality tiers or types (in this case, flat or stylistic graphic design artwork)?

I challenged AI ImageGen with “Flat vector of two people building a house, infographic.”

I was quite surprised—it created perfectly serviceable graphics. I could easily see using these in a blog post, presentation, or social media update without anyone questioning that it was produced or processed with digital tools with high production capacity like AI rather than individually drawn manually with human-created resources. (Yes, I know there are some errors.)

After experimenting with AI ImageGen, I have a good feeling about this tool. It produces realistic content across different styles—something that many alternatives I’ve used do not do.