- GPT-4o
- Midjourney
GPT-4o.
Remember when everyone was generating Studio Ghibli–style AI images of themselves, their pets, and anything else they could think of? That was all powered by the built-in image generation capabilities in ChatGPT -4o
It can perform both text-to-image and image-to-image conversions for every style of art, making it helpful for a wide range of applications. I use it myself to generate royalty-free images to put on blog posts and social media posts—and now that it can do reliably text inside of images (something that tools like DALL·E 3 couldn’t quite do), it’s even more useful.
The ability is on the free plan of ChatGPT, albeit with daily usage limitations. You can lift those by moving to a Plus or Pro plan.
Midjourney
Midjourney is the OG of AI image creation. It’s been around since early 2022 and is still widely popular because of its painterly quality and for creating source images to generate a video from an image.
Honestly, I pretty much exclusively use the native image features in my go-to AI helpers like ChatGPT or Gemini. But when I’m looking to create something truly lovely, Midjourney is what I turn to.
The only disadvantage is that Midjourney now does not provide a free plan—you will have to subscribe to one of its paid tiers just to create a single image. There are four plans to choose from, ranging from $10/month to more expensive tiers that provide more generation time and features such as video support and stealth mode.