The best AI image generators available right now

From Pope Francis in a puffer jacket to Wes Anderson-inspired movie trailers, AI-generated images have become increasingly common on the Internet. It’s not hard to see why — modern AI tools have become so convincing that their images have even won prestigious photography awards. You don’t have to be a talented artist or photographer to generate beautiful art either. If you’ve ever used a chatbot like ChatGPT or Bing Chat, you’ll feel right at home using any one of the best AI image generators.

In order to generate an AI image, all you need is a text prompt that describes the scene you have in mind. For example, I used the prompt “Golden hour New York City skyline with towering skyscrapers and bustling streets” in one of the images below. Your imagination can go much wilder, though, with detailed prompts that specify each and every intricate detail.

So with those capabilities in mind, here’s our list of the best AI image generators currently available. We’ll keep this list updated as technology continues to evolve.

The best AI image generators

From free to paid and simple to complex, you’ll quickly learn no two AI image generators work the same. In fact, almost all of them differ in terms of their underlying machine learning models too. This means that identical prompts will give you different results depending on which generator you use. To that end, here are your options:

While Midjourney wasn’t the first AI image generator, it has now become the go-to choice for many. This is because the latest V5 iteration of Midjourney has overtaken most of the competition. In some cases, the results it produces can look photorealistic, albeit with some minor mistakes that reveal its AI heritage.

Unlike most other services on this list, you can’t access Midjourney through a website or app. Instead, you’ll need to use Discord — the chat app commonly used for gaming communities. That’s not a bad thing, though, as Discord runs on all major platforms, including a web browser. To get started, simply create an account and join the Midjourney Discord server. You may have to accept the terms of service before getting started, but that’s the only barrier to entry.

New members only get to generate a limited amount of free images (around 25) as part of a free trial. However, Midjourney disables the free trial during periods of heavy traffic. The cheapest way to use the service is a $10 monthly subscription. You can save a bit more if you commit to the subscription for a full year. Unfortunately, this tier still only grants you roughly 200 images per month.

While you can use Midjourney with basic prompts that include just a few words (or even an emoji), its real power lies in parameters. For example, you can specify a 16:9 aspect ratio using the --ar parameter or create anime-style images with --niji.

Dall-E became the first generative AI to break into the mainstream when it launched in 2021. The latest version, Dall-E 2, takes things to another level with better language understanding capabilities and higher-quality images. It can also edit existing images, allowing you to replace or add entirely new objects.

OpenAI, the company behind ChatGPT, created Dall-E. The company used its GPT-3 large language model as a foundation, which explains how the image generator understands your prompts.

Using Dall-E is quite simple and straightforward. You’ll need an OpenAI account, which you may already have if you’ve ever used ChatGPT. Once logged in, you can enter your prompt into the text box and hit the Generate button.

Each image you generate will cost one credit. Unfortunately, however, Dall-E no longer gives new users any way to obtain free credits. That practice ended in April 2023, so if you create an account today, you’ll need to buy credits. You’ll have to load $15 each time, which results in 115 credits. That works out to roughly 13 cents per use, which isn’t too bad for one of the best AI image generators out there.

Having said that, you can still use Dall-E for free through Microsoft’s Bing Chat. Scroll down for more on how to do that.

Unlike the other AI image generators on this list, Stable Diffusion is completely free and open-source. If you own a powerful computer with a dedicated graphics card, you can even download and run the model yourself. Even better, it works completely offline. However, this is not an option for everyone as the minimum requirements call for a GPU with 8GB of VRAM. That disqualifies most non-gaming or non-workstation computers.

The good news? You can use Stable Diffusion to generate images entirely in your web browser. Just visit the StableDiffusionOnline website and type in a prompt. It doesn’t get much simpler than this.

You don’t need an account, but your results will be saved and visible to others. For maximum privacy, you’ll have to run the model on your own computer. You can’t see your personal generation history either, so it’s only meant for one-off usage. Having said that, the platform does let you browse through past prompts and their results. This can come in handy when you’re lacking artistic inspiration.

While StableDiffusionOnline from the above section works well enough for a free service, it’s not very feature-complete. You can’t customize the size of the image, for example, nor can you edit them after generation. But luckily, the creators of Stable Diffusion, Stability AI, have just the solution in the form of DreamStudio.

As soon as you open DreamStudio, you’ll notice that it looks like a polished app with many buttons and dials. You can generate multiple images at once (or fewer), change the aspect ratio, omit certain elements in your AI-generated masterpiece, and more. DreamStudio also offers different “styles”, covering everything from photorealism to origami and comic book looks.

On top of all that, you can also add and remove objects in your AI-generated images using the edit tab. For example, try using the eraser tool to remove certain parts of the image. Then type in a prompt to fill up the blank space.

As an official product and a highly polished one at that, DreamStudio does cost money to use. But the good news is that you get 25 free credits when you sign up, which will get you 125 generations. Beyond that, the minimum $10 purchase grants you 1,000 tokens or roughly 5,000 AI-generated images.

Microsoft’s Bing search engine gained a powerful new chat mode shortly after ChatGPT’s release. Since then, it has only become more capable with the company adding a free AI image generator into the mix as well. The service, dubbed Bing Image Creator, has a dedicated page now with ideas and instructions on how to use it.

Since Microsoft works closely with ChatGPT creator OpenAI, it’s no surprise that Bing’s Image Creator uses Dall-E under the hood. That means if you run out of free credits on one platform, you can simply switch to another.

When you sign in with your Microsoft account, the Image Creator automatically grants you 100 “boosts”. These are essentially credits that allow you to generate a new image nearly instantly. Boosts replenish every week automatically, and you cannot pay for more credits at this point. What happens when you run out of boosts? As the name hints, you’ll have to wait much longer. But the good news is that you’ll still be able to generate an unlimited number of images for free.

You can also access the image creator through vanilla Bing Chat, even on mobile. Simply switch over to Creative mode and start your prompt with “Generate an image of…” This method follows the same limit as the chatbot for now, at 200 conversation threads per day.

That’s it for our list of the best AI image generators available today. Have you used any of them yet? Let us know which one you’d use over the competition in the poll below.