AI image generators have taken the world by storm in the last year or so, raising moral questions alongside questions surrounding copyright of said images. They're undoubtedly useful tools for anyone who wants to make images for their own personal usage, such as for a Dungeons and Dragons game, but anything past personal usage instantly gets murky.

Having said that, these are some of the best AI image generators available on the market today. Keep in mind that their use in commercial contexts hasn't really been tested, and we suggest only using them for your own personal projects. With it unclear how lawsuits surrounding their training data will resolve, as many artists feel that their images were unfairly used in the training data, things could rapidly change in this sector very quickly.

Stable Diffusion

The best and easiest

Stable Diffusion is my favorite AI image generator, for the simple reason that you are in total control of it. It runs on your computer, and you control the code and the models that you use, and you can even train it on your own face if you wish. There are web GUIs that you can download and set up, but you'll need a relatively fast computer to generate images. It's totally free and you control everything, but the downside is that controlling everything means you need the hardware to run it, too.

AI image generation is particularly vRAM heavy, so you'll need a good GPU, and these tools are all made with Nvidia GPUs in mind. Stable Diffusion can also do things like upscaling images and img2img, which will take a basic drawing you make and turn it into a high-quality image.

DALL-E 3

You can use this one for free

DALL-E 3 is the latest DALL-E image generation algorithm, and it's made by OpenAI. You'll get it for free with Microsoft Copilot, but you'll also get it if you pay for ChatGPT Plus. It's capable of rendering images just like Stable Diffusion can, but you don't need any beefy hardware to do it. Also, it's significantly better at text than any of its predecessors in the industry, which makes it a lot better for generating images that contain text somewhere.

• ChatGPT

  ChatGPT is one of the best LLMs out there, and it's completely free to use. You'll need to make an account, but there's not much else to it. 

  See at ChatGPT

• Microsoft Copilot

  Copilot is an artificial intelligence chatbot available for iOS and Android that uses the DALL-E 3 and GPT-4 models. The software is also built into Windows and can be accessed over the web.

  See at Google Play Store See at Apple App Store

Midjourney

No longer free

Midjourney was free for quite a while through a Discord server, but now it costs money to use. Starting at $10 a month, you'll be able to generate images that take up to 3.3 hours of GPU time per month. That's not bad considering images will be generated in less than a minute most of the time, but keep in mind that Copilot and Stable Diffusion both offer free options.

Google's ImageFX

Included with Gemini

You can use Google's ImageFX with Gemini, a free image generation tool powered by Imagen 2. In my opinion, it's not as powerful as what you get with Copilot or ChatGPT, but it gets the job done if you want to generate images within Google's ecosystem. Google's Gemini also gets you a lot with an Advanced subscription, more than ChatGPT Plus does, anyway.

These are some of the best mainstream image generators

Some free, some not

If you're looking to use an AI image generator, these are some of the best in the business. I personally recommend using Copilot, especially given that DALL-E 3 is free. If you want to generate an unlimited number of images and have a powerful PC though, I highly recommend Stable Diffusion. It's easy to use, quick to setup, and highly customizable.