The Best Free AI Art Generators, Ranked (2024)

What is AI art? Well, of all the many, many ways armchair philosophers have come to define art and the act of human expression, AI-generated images don’t have much in relation. There is certainly skill in knowing the correct type of prompt to get the results you want, but there’s no real intent behind an image. If the AI is the “artist,” how can any human claim the generated output as their own?

The question then becomes, what use is AI-generated images? This artificial art has become a shortcut for lazy PR professionals asked to pump out ill-fitting images that don’t hold up to even cursory scrutiny. Entrenched art communities from sites like DeviantArt have largely decried any and all AI-generated images.

Now you have companies like Shutterstock and Adobe with their in-built AI image generators, there’s a new slate of services that facilitates AI art without any of the concern over copyright. Getty Images also has its own AI art generator built into its UI. Yet, it’s still suing the makers of Stable Diffusion, Stability AI, for using copyrighted photos to train its art generator. Artists are also looking to sue some of the makers of these AI generators, but so far they’ve had difficulty making their lawsuits stick. So for all intents and purposes, the legal identify of paid-for AI images is still in flux.

In that way, I can only advise that you keep these images for your own use and enjoyment. That’s why it’s better to stress the “free” part of these rankings, as you’re not going to have anything like a professional standard image quality.

And even then, there really isn’t much “free” about these AI models. Like any good dealer, the first hit is always free. These sites usually offer a few tokens to generate a few pictures before inevitably asking users to pay for more images or to upscale those pictures. You could look at one of the most popular AI image generators, Midjourney, though that service is no longer free even though it started as a free Discord server back in 2022.

There’s also a wave of AI video generators begging for attention. So far, OpenAI’s Sora model seems like the most capable of all that’s come so far. Even the strangest, most-off putting video from this model is leagues beyond what’s currently available from services like Runway. Unfortunately, it’s also locked behind closed doors, so we may not get access to it for a while now. Microsoft also has its own video generator called VASA-1, which can create some pretty lifelike deepfakes except for their exceptionally weird teeth.

Some models like DALL-E 3 are restricted to those who pay for a ChatGPT subscription, though thankfully Bing’s image generator, now called Image Creator from Microsoft Designer, uses the same model, and that’s available for free. You could go back to DALL-E 2 instead, but its not as capable as modern models.

For the purpose of these rankings, I wanted to ignore the hubbub around total terabytes of training images. Still, the total time it takes to create each image, their standard free resolution, and usability are all taken into account. To best rank each program, I gave them all the same, rather esoteric, text prompts based on some books I’ve recently read. Those books include:

Under the Pendulum Sun by Jeanette Ng

Prompt: “A man and woman stand under a pendulum sun in the heart of Arcadia.”

The Dispossessed by Ursula K. LeGuin

Prompt: “A lone mathematician stands on a dusty planet owning nothing.”

A Memory Called Empire by Arkady Martine

Prompt: “A foreign woman struggles alone against the machinations of a cosmic empire.”

There are several image generators that require photos instead of text prompts. As much as I would like to keep it consistent, I want to be inclusive of different systems rather than exclude them. For the image-generating platforms that don’t allow for text prompts, I used the same image for each one:

There is practically no legal precedent for AI-generated images. It’s a well-known fact that these AI models are trained using many gigabytes of copyrighted art and photos scraped from the web. Still, until any of the manylegalcases actually sets a roadmap for using any of these generators, we suggest you don’t try and use these free AI art generators for anything but a fun diversion.

And that’s why I find the weirder the generated output, the better. There’s a whole lot of weirdness with AI-generated video. Gizmodo has its own list of the strangest AI-generated commercials you can find here. With the advent of text-to-video models, there’s even more potential for off-the-wall generated content, so we’re including the free video generators in our rankings.

I’m no art critic by any imagination. Still, at least I can tell whether AI-generated art attempted to depict a prompt in a way that’s not derivative or relies upon copying and replicating art found on the internet.

Want to know more about AI, chatbots, and the future of machine learning? Check out our full coverage of artificial intelligence, or browse our guides on How to Use ChatGPT and Everything We Know About the OpenAI chatbot.

Advertisem*nt

2 / 27

The first text-to-video AI generator to catch the internet’s attention, ModelScope, was released in March 2023. The AI video generator is rudimentary, producing only 2-second clips that feature odd distortions, much like early AI image generators. Many of its videos also bear the Shutterstock logo, hinting at where it gathers its training data from.

From Gizmodo’s test drive of ModelScope:

The AI text to video system called ModelScope was released March 18th and already caused some buzz for its occasionally awkward and often insane 2-second video clips. The DAMO Vision Intelligence Lab, a research division of e-commerce giant Alibaba, created the system as a kind of public test case. The system uses a pretty basic diffusion model to create its videos, according to the company’s page describing its AI model.

Advertisem*nt

3 / 27

There are a few new models being created off the ModelScope prototype, including zeroscope_v2, a family of open-source models. This creates higher-quality video than ModelScope, and the XL version can upscale it to 1024x576 resolution. It’s available on HuggingFace.

Generating video from the model can take quite a long time if you’re using somebody else’s public space, but the results are pretty interesting just by themselves.

Despite advertising as much more capable than others, text-to-video is still in its very early stages. Give it a simple prompt like “Man walking through the forest,” and you’ll receive a rather simple rendition of what you asked. Give it anything more esoteric or imprecise, and it will spit out some pretty wild visuals. To be clear, I prefer some more stylistically strange, but people morphing into each other isn’t quite what I intended with a prompt.

Advertisem*nt

4 / 27

Runway.Ml’s much-hyped text-to-video generator is the best version you can get, but that’s more indicative of how far these generators need to go to create anything anywhere near as capable as most other AI image generators.

Most videos include a simple pan or dolly shot while the subjects on screen amble about or stare off into the distance. The picture quality is very similar to Runway’s text-to-image generator (which is the precursor to Stable Diffusion if you think the video above looks similar to different slides).

The benefit of Runway’s model is you can make the video go for more than 10 seconds, whereas the other models are far more restricted. Users receive just over 100 seconds of video generation before the company asks them to sign up for one of the company’s paid subscriptions.

While Runway produces a much bigger quality video than the other two models can, it’s not anything more than a cute gimmick. Simply put, text-to-video is cool and fun, but it has a long way to go.

Advertisem*nt

5 / 27

Advertisem*nt

6 / 27

Advertisem*nt

7 / 27

The system, once known as Dall-E Mini, now Craiyon, was originally supposed to offer a quick, free AI image generator to tide people over until the full release of OpenAI’s DALL-E 2. The model has received a few updates more than a year since its release, though it’s much more of a money-upfront kind of business. You can generate a few images of varying quality but are very low-res. To perform any upscaling, you need to pay up. If you don’t want a watermark, you’ll also need to fork over some cash.

The model releases a fair few outputs, but none are especially nice to look at.

Advertisem*nt

8 / 27

Shutterstock has dived head-first into AI-generated content, unlike some of its stock service contemporaries. The company started facilitating the sale of AI content late last year and inked a deal with DALL-E creator OpenAI. To access the generator, users need to sign up for Shutterstock. While you can use the service to download images, those without a subscription still need to pay for each generated image, so it’s hard to call it truly “free.” So, despite the images being somewhat better quality than others, that’s why it’s ending higher up on the list.

Shutterstock released its DALL-E 2-based AI image generator last year. The service generates four images at about 500x500 pixels, which is pretty sizable compared to some of its competing platforms. Users can set the output between five separate “styles” to make the generated image look “3D” or more like a digital photo.

Like all the other generators in this list, I used the most basic output possible and received a few surprising results. The system generated several rather interesting Under the Pendulum Sun-inspired images, though like all AI image generators, it clearly has problems with straight lines and latticework. The system took obvious inspiration from picture book style for its multiple renditions of A Memory Called Empire. It could almost be considered artistic, save for the misplaced arm and awkward shadow running along the woman’s back.

Another consideration is how Shutterstock claims it is using contributor’s photos and images to train the AI. The company has promised to compensate contributors whose images train the AI through a so-called “Contributor Fund.” Image contributors receive a “share” of the pie based on how many images they’ve uploaded to the site.

Advertisem*nt

9 / 27

Hotpot.ai’s art generator is a pretty rudimentary program, but it has become much more capable since I first tried it. The company is also advertising different features, like tools to remove objects and backgrounds from photos, though those do indeed cost money.

The whole “free” aspect is where Hotpot hurts the most. Creating small, thumbnail-sized images is free, but doing anything more, including resizing or allowing “commercial use,” costs a subscription. I’m surprised by the quality of the images generated, especially how each image definitely has a lot of classic Sci-Fi and fantasy feel to it, but the limitations on the free versions hurt it the most.

Advertisem*nt

10 / 27

Runway ML was one of the two startups that helped give us Stable Diffusion, and its free text-to-image service has the same quality as the earlier open-source versions of the software. According to Forbes, as Runway continues to get millions in investment funding, the company has moved on to AI-based video editing tools and image expansion services, along with the rest of its suite of customizable AI tools, most of which require payment to get the most out of them.

But for this free service, users get just 25 images to start before being asked to upgrade, which isn’t so great compared to others. When using these non-specific prompts, the actual quality of images leaves quite a lot to be desired. Also, like Stable Diffusion, Runway has a fascination with fake text meant to resemble a book cover.

Advertisem*nt

11 / 27

A lot’s been said about Stability AI’s free art generator while it was in closed beta, but on Aug. 22, the free image generator finally got a full release, and since then, it’s been one of the more talked about image creators. Stable Diffusion is open source, free, and unfiltered compared to the likes of Dall-E or Google’s (still publicly unavailable) Imagen. Its basic page on Hugging Face is pretty rudimentary, but you can make it much easier to use with some simple freeware tools, though you’ll have to install Python to get it working. Stable Diffusion relies on a model based on the LAION-5B data set that filters out watermarked images and logos, according to its own page. Based on a report from tech blogger Andy Baio, many of the images are mostly sourced from Pinterest and other photo and art blogs. My attempts at asking it to create its own art without telling it to ape copy one particular artist were largely unsuccessful. It kept offering me black-and-white images without any style or substance. Without asking it to copy a specific artist, it only offers pretty disappointing works compared to what other users have managed to get.

It seems the system is much better at coming up with art when you specifically ask it to do it in the style of a particular artist. This, of course, introduces a host of ethical problems, especially for the living artists people may be emulating, as shown by fantasy artist Greg Rutkowski, who was interviewed by MIT Technology Review and said he was worried about the number of fake AI art bearing his name would eclipse his own visibility.

Advertisem*nt

12 / 27

You got to love freeware, and this Python-based Google Colab notebook is relatively easy to use, so really, you have to give major props for a system that’s both relatively simple and open and available to all users. All you have to do is go to the link, scroll down, and input your prompt in the text box, then either hit Ctrl+F9 or Runtime - Run All.

But of course, we’re ranking system’s artistic ability. The system progressively iterates on the design so you can see where the AI is trying to go with each image. My prompts offered some really interesting results but a few head-scratchers as well.

Advertisem*nt

13 / 27

The Wombo Dream model has gone through several iterations, starting as a GAN model before using the standard diffusion model, and honestly its far more boring because of it. You can still choose to use a variety of different styles, but most of them are locked behind a premium subscription. You can generate one image per prompt for free, but to get four per prompt you need to pay $10 a month or $90 a year.

The actual images aren’t very creative either, nor very detailed when using the default Dreamland V2 style. I received much more creative images on different styles, which probably goes to show how much style it needs to borrow from the different artists in its training data. They also all come by default in a “trading card” format with a tall aspect ratio that seems good for creating phone backgrounds and not much else.

Advertisem*nt

14 / 27

StarryAI has updated its model from a GAN-based art generator to a much more recent diffusion model, and yet it is somehow less interesting because of it. The images are upscaled far more than they were before, but instead of the manic shapes of before, the new one can generate derivative works that, while looking interesting enough on their own, don’t actually have much going for them. The AI art generator has a rather clean web-based interface that lets you create images in a few styles, but the regular StyleVisionXL offers a rather generic look. Based on the images alone, it uses a version of Stable Diffusion, so don’t expect something too far out of the norm.

You receive a meager few credits daily, which you can use to generate four prompt versions. Again, without adding details to each prompt and requesting a specific style, the generated image doesn’t have much going for it.

Advertisem*nt

15 / 27

Google’s Deep Dream Generator claims it can transform images into stylized works of art. The main feed of images on the site does make it seem very evocative, but of course, it’s not exactly easy to make an image of young Skip appear as more than just a dog with a Photoshop filter on.

I put young Skip in the style of Pieter Bruegel, the famed Dutch artist behind works like The Blind Leading the Blind. It came out… okay. I tried it with a mandala pattern, and it was… interesting. I then added a Salvador Dali painting to the style, and it came out boring.

It’s a fun tool, but I feel like I’m playing with a sophisticated color-in-the-lines book rather than generating any real art.

Advertisem*nt

16 / 27

The Nightcafe system will only let you do a few images before asking you to pay up, but like some other AI image generators, it gives users a whole selection of different styles to choose from. It wants users to buy “credits” in order to make more art or bump up the quality and resolution of each image, and you will run out fairly quickly.

The art itself is all over the place. I appreciate what it tried to do with The Dispossessed, but it wouldn’t even create a human-looking being in either of the other two prompts. The image resolution is also not great, and the art is very strange, even in the default oil painting setting. Sure, it’s nice to be able to try out the system to see if you like it first, but even for just $10 a month, I can’t say you won’t find better options out there.

Advertisem*nt

17 / 27

DALL-E 2 may have been the most hotly anticipated AI image generator since it was initially put into its beta release back in April. However, as shown by these examples, there’s been multiple releases to cash in on the AI-art craze.

OpenAI, the makers of DALL-E, finally released their system to the wider public on Sept. 28. Like many other systems, it offers users a slate of 50 free prompts, plus a few more each month. Users can also pay for more image generation credits.

My experiences with DALL-E have been interesting, to say the least. I find the system does not compute specific artists too well, and even when you give it pretty detailed instructions, it defaults to an impressionistic art style. Of course, that’s not what this ranking is looking to do. For this, I’m more interested in what each AI system is capable of on its own without giving it an image to specifically replicate. DALL-E’s systems obviously draw from a whole host of real photos and art, but they still largely fail to replicate faces. However, the generator did seem fond of the prompt for A Memory Called Empire, and it offered several inspiring images that captured the essence of the book.

DALL-E 3 has been out for about a year, but as of now, the only way to access it is by paying for ChatGPT Premium. However, that also leaves Bing Image Creator, which also uses a version of DAll-E 3.

Advertisem*nt

18 / 27

The DreamStudio AI art generator is free, and it has a lot of bells and whistles other AI art generators don’t have, like the ability to scale the width and height of the image while telling it how close you want the image to be to your prompt. It’s developed by Stability AI, and it uses the later version of Stable Diffusion. It essentially works as an easy-to-use platform for Stable Diffusion, though it uses a credit system that will eventually demand payment.

The system is also surprisingly fast, just by the examples above you can see the AI isn’t afraid to mix and match art styles on a whim. You start with 100 credits, and depending on the image size and number of generations each prompt might cost you more or less credits. You start with 100 credits, and $1 equals 100 credits.

Many of the outcomes were very reminiscent of other Stable Diffusion creations, especially the ones that emphasize the cell shaded or comic book style. Still, it’s a nice system if you just want to bang out a few quick images with little fuss.

Advertisem*nt

19 / 27

Google released ImageFx in February this year alongside its MusicFX and TextFX models as a way to gin up hype for its AI initiatives invading Android and its load of other apps. It’s all available in Labs, which is Google’s experimental branch of services. You have to sign up to gain access, and space depends on Google’s whims.

Still, if you manage to get access, ImageFx is one of the better free services available simply because there’s no credit system, and you can create as many images as you might like. It also offers synonyms for certain nouns or verbs, though maybe not in many ways that are very helpful. Unfortunately, the system is also very restrictive in what you might use as your prompt. Some of the generated images are quite appealing, while others are very, very bland.

Advertisem*nt

20 / 27

Adobe Firefly Image 2 is an interesting beast. For one, the web-based application has a fully free plan that offers 25 credits every month. That’s pretty good compared to other services online. Two, it uses the company’s AI model that only uses images from its stock site, for better or worse.

The actual outputs feel very basic in that way. Turning it to the “art” content type leads to a very ad-friendly kind of output. It’s not enough to wow your friends, but the site itself offers a fair bit so long as you sign up for Adobe’s site.

Advertisem*nt

21 / 27

Stable Diffusion XL was Stability AI’s attempt to really cement itself in the rapidly expanding market of image generators made for businesses. It was supposed to be better than Stable Diffusion 2.0, the sequel to the first Stable Diffusion. It’s not fully available online, but you can generate images for free with the “Try SDXL” button on the company’s website. The code is also available if you have the capabilities and know how (or the cash to run it on one of Hugging Face’s servers) to run it for free.

Like older SD models, it’s still a little loose in the face and hands departments when you don’t wish to provide specific instructions in your prompt. That being said, some of the images are fairly interesting and they are a big step up from earlier images created with the same prompts.

Advertisem*nt

22 / 27

Microsoft’s partnership with OpenAI has brought us ChatGPT in Bing and Bing in ChatGPT. Still, it’s also allowed anybody with a Microsoft account to access 99 free prompts per day for AI-generated images. The site says that if you run out of prompts, then the system will take longer.

Image Creator, which was previously dubbed Bing Image Creator, is essentially the DALL-E 3 model found inside the Bing webpage, so it’s already plainly capable. Each prompt creates four 1024 × 1024 images, which’re clearly just as advanced as OpenAI’s product. What makes the system better is how many prompts you get for free.

Not only that, but it creates some interesting interpretations of my strange prompts. Of course, it’s still Microsoft, and the most annoying part of the tool is the obvious watermark at the bottom of every generated image.

Advertisem*nt

23 / 27

CatBird is another AI image generator site to hit the scene after the big late 2022 boom, and it’s already one of the strongest contenders, especially as so many other services have limited their free options.

It’s not running any specific generative AI model, but a slew of them at one time, including Openjourney, Dreamlike Diffusion, and Stable Diffusion, among others. They’re organized based on how photorealistic, anime-esque, or fantasy-styled they are. By default, the system bases its output on several models based on a single prompt, and it creates three images at a time in a 512 by 512-pixel ratio, though upscaling those images is currently free.

The images themselves are very enticing, especially as you get a wide range of styles from each. The prompts also auto-generate words like “best quality, high quality” to try and give users superior results. Eventually, the service will start offering a “Pro” paid version that will speed up wait times and offer six images at a time and more models per prompt. The system also does not have the ability for accounts, but CatBird did tell users on its Discord that the company is working on a feature for users to save their images so they can upscale them later.

Advertisem*nt

24 / 27

Leonardo’s main page includes a spread of separate AI image models, but the main Leonardo Diffusion tool has a sizable suite of options for creating different images. You can set your resolution size for the generated images and decide whether to use the tool’s extra bells and whistles that supposedly up the contrast on each new picture.

Of course, enabling more settings and upscaling images will cost more tokens. You receive 150 to start, but those can run out quickly if you want the best-quality pictures. Subscriptions start at $12 a month, but there’s a discounted rate if you sign up for the year.

The images themselves are pretty stunning. They’re easily some of the most stylized of created images next to Midjourney, but unfortunately, there’s even less information immediately available about what kind of models Leonardo uses. Users can even train their own model with different images that can fine-tune results, but that’s not what this ranking is about.

Advertisem*nt

25 / 27

Midjourney is easily one of the best AI art generators despite being tied to its own Discord server. It used to be free, but unfortunately, since last year, the service has now required a subscription. It’s a pain point for those who used to get their kicks by hopping onto the Midjourney chat to see what the hell folks on the internet were trying to make the system generate next. One day, it may return to the fold by offering a free version. Until then, there are plenty of fine, cheap options to go around.

Eventually, Midjourney will open up its site

Advertisem*nt

26 / 27

The company behind our favorite AI art generator is launching a monthly magazine. A subscription will cost you $4, though the company is offering copies of the first issue for free.

Why would a potential subscriber want to pay for print copies of free online images?

“We thought it would be fun, and the community agreed, so we did it,” Midjourney founder David Holz told Gizmodo.

More from Gizmodo’s story on Midjourney magazine:

The company behind the tool will feature images in its simply and uncreatively titled publication picked from among the 10,000 images rated highest by its site’s community members, according to its website. Interviews with image makers and Midjourney enthusiasts will also appear in the magazine. The Midjourney Community Showcase page displays images submitted and rated by users.

Want to know more about AI, chatbots, and the future of machine learning? Check out our full coverage of artificial intelligence, or browse our guides on How to Use ChatGPT and Everything We Know About the OpenAI chatbot.

Advertisem*nt