On today’s Roundup, we’re taking a look at every AI image generator I could get my hands on and putting them in their own tier list!
Drop us your thoughts at news@mspmedia.tv.
Story Links:
Make your own version of this Tier List:
https://tiermaker.com/create/ai-image-generators-tier-list-16259594-2?ref=list-remix
Adobe Firefly/Generative Tools: A TIER
Requires an Adobe subscription starting at $35/mth for unlimited generations and access to Firely Generative AI Tools in Photoshop.
Touts one of the most responsible data training philosophies, even though they have recently been called out for supposedly using data without artist consent, they responded promptly and removed said images.
Craiyon: C TIER
Free-to-use, ad supported platform and one of the original image generators to really go viral. Craiyon has made massive improvements since it first came out on the scene as Dall-e Mini and really stole the thunder from OpenAI when Dalle 2 was mostly only available in a very limited beta. It’s not SD based and claims they use their own model that looks very similar to Dalle’s earlier creations and while the images don’t look great, Craiyon will always have a special place in my heart as my first ai generator.
Dall-E 2: B TIER
Reasonably priced, especially if you can just be happy with free credits, images are pretty janky looking. Moderation is not as tight as some other platforms.
$15 for 115 credits, 50 credits initially, with an additional set of 15 credits added every month.
Dall-E 3 via Bing Image Creator: S TIER
https://www.bing.com/images/create
Very high quality images with the major perk of being able to generate text better than any other generator I’ve used. Pricing is fair as Microsoft gives 25 free boosts credits per day with images simply taking longer to generate after you’re out of boosts. You’re able to purchase more boosts with Microsoft Rewards points. Moderation is quite tight and very inconsistent with no recourse for a blocked prompt plus threats of account suspension.
Deep AI (SD Based): D TIER
Free-to-use with paid features. Images were pretty standard SD quality.
Pricing: $4.99/month for 500 generations. Moderation was strict with block words but apparently allows “Unfiltered Content” for paid users.
Deep Dream Generator (SD Based): D TIER
https://deepdreamgenerator.com/generator
Free to use with 100 credits to start and 5 credits per generation. Starting at $19 with 120 credits and a new credit regenerating every hour.
Interface is easy to use, it’s basically a very basic webui for SD as far as I can tell. The moderation is extremely slack. I had no prompts blocked.
Dream by Wombo (SD Based): C TIER
Pricing: Free version available (with ads and restricted features), Premium at $9.99/month.
Known for abstract design and easy usability.
Fulljourney: A TIER
Free to use on the Fulljourney Discord server. Capable of much more than just image generations. Pricing is quite steep but the quality of the generations for images is quite a bit higher than most SD images but not quite on the MJ or Dalle 3 level.
The options for purchasing credits are:
• Starter $7.99 USD 40 credits
• Hobbyist $19.99 USD 100 credits
• Creator $31.99 USD 160 credits
• Legend $99.99 USD 400 credits
Jasper Art: D TIER
Too expensive, surprisingly difficult to use, and quality of images were on par with Stable Diffusion. Notably, the faces looked strange. Moderation is tight but it is at least using context of the prompt rather than a list of banned terms.
Pricing: Creator Plan at $49/mo, Teams Plan at $125/mo. No free image generation credits.
Leonardo (SD Based): A TIER
Highly customizable and high quality image generation tool.
Free-to-use with 150 credits (tokens) to start and then $10 for 8500 credits (tokens). Changing settings such as resolution and aspect ratio affect the amount of tokens required per generation. Leonardo is an exciting platform because the creators are iterating far beyond the basic SD features. They offer some of their own workflows, one of the newest called Alchemy which has some complex processes for increasing image quality and detail.
Midjourney: S TIER
Quite pricey with currently no free options. Quality of images is extremely high, notably hands and faces almost always look correct. Medium difficulty due to having to use Discord like a terminal window to operate. Moderation is tight but it uses context of the prompt rather than a list of banned terms. You also have the opportunity to appeal a blocked prompt which has worked most of the time for me.
Pricing: Basic Plan at $10/month, Standard Plan at $30/month, and Pro Plan at $60/month.
Nightcafe (SD Based): B TIER
https://creator.nightcafe.studio/studio
Free to use with 5 free credits per day. Image generations range from 1 to 1.5 depending on the image settings. Usability is great but very similar to SD as it basically is SD with some other legacy features from their older interfaces built in.
Best for generating creative images.
Pricing: Free version available (with limitations), plans starting at $5.99/month for 100 credits
Shutterstock: B TIER (Dalle powered)
https://www.shutterstock.com/ai-image-generator
Pricing starts at $29/mo. Touts a similar “ethically based model” as Adobe Firefly based only on their stock photography library. But the image quality was nothing to get excited about. The interface is extremely basic with only the option to type a prompt and add art style modifiers. Way overpriced unless you are already in need of shutterstock’s photos, then this is a nice perk.
Stable Diffusion: S TIER
Free to use and open source! The only cost is the energy your computer uses or the cost of a virtual server for hosting if your computer is not powerful enough. Lots of options for interfacing with SD from using a terminal window to many free open source WebUIs like Automatic1111 or ComfyUI and many more. Image quality is medium to high depending on the models you use. And this goes a lot further once you add in things like ControlNet, Extensions, and Scripts. The biggest downside of SD is that there’s a pretty steep learning curve to get the most benefits out of it. You can of course use stable diffusion in so many of the sites we will or have discussed, but the capabilities are expanded exponentially when you are using a customized interface with additional tools like trained models, Loras, Hypernetworks, Lycoris and more.
Starry AI: (SD based) C TIER
Free to use with a limit of 5 artworks daily.
8.99/mo $108/year
200 Image generations per month
31.99/mo $384/year
3000 Fast generations per month
To the best of my knowledge this looks to be a stable diffusion based service, so hard to justify the cost if you’re willing to invest some time in learning SD.
Medium quality images, depending on what you’re trying to create. Faces and hands are not great. Text ok.
SynthesysX (most likely SD Based): C TIER
Good that it’s free and very easy to use, but applications are limited and Premium plans seem very not worth it. This is basically an applied use of Stable Diffusion’s image-to-image tool built into a Chrome browser.
This is a Chrome extension that you can use by right clicking on images in your browser and “Revisualizing.”
Free for 50 images a month, and premium plans start at $12 a month for 200 generations.
Connect with our host by joining the MSP Media Discord:
Be sure to follow us on social media:
Facebook: https://www.facebook.com/mspmediatv/
Twitter: https://twitter.com/mspmediatv
LinkedIn: https://www.linkedin.com/company/mspmediatv/
Instagram: https://www.instagram.com/mspmediatv