I’m teaching Gen AI to 35 teenagers tomorrow, and I was thinking about which AI Art engines I should introduce them to.
I tried the same prompt – “A quick brown fox jumps over the lazy dog” – on six different engines to compare results.
The free imaging engines are Bing Image Creator (aka Microsoft Copilot Image Creator), Google Gemini, and Meta AI. The paid ones are ChatGPT Plus ($20/mth), Midjourney ($10/mth) and Adobe Firefly, part of the Adobe Creative Cloud suite ($20-50/mth).
FREE APPS:
I like Bing Image Creator the most as it follows prompts well, and the image quality is often good. However, it only churns out square images. It uses the DALL-E 3 engine found in ChatGPT Plus.
Google Gemini and Meta AI are hit-or-miss. Sometimes, you get very nice visuals; other times, you get janky images. The visuals are also locked to square aspect ratios.
If you don’t want to pay for AI images, I recommend using three engines and seeing what you get. I’ve tried other free apps like Freepik and DeepAI, but they failed terribly with my text prompt.
PAID APPS:
Midjourney still produces the best visual quality, but it doesn’t follow prompts very well. What it really excels at is generating a diverse range of art styles that can inspire you to think in new directions.
Adobe Firefly and Photoshop are a few steps behind in visual quality, but they have a much better UI that lets you select different artistic styles. They can also generate visuals that closely reference your uploaded images. This is a godsend for anyone who wants to create beautifully finished visuals from quick sketches.
ChatGPT Plus is the weakest of the paid apps for image generation. Although it uses the excellent DALL-E 3 engine, it only generates one image at a time (the others often give you four choices), and the images often look generic. However, it will make landscape or portrait images for you.