The Best AI Art Websites & Generators (Free & Paid) w Rankings Guide for 2024

AI art image generation it's absolutely blowing up lately it's everywhere there are literally hundreds of websites where you can go generate with AI art but how do you know which ones are actually worth it how do you know which actual AI models are the best it can get very confusing well thankfully I am here for you today this video is going to be a massive deep dive into the AI art generation world and you'll notice there's chapters where you can skip around looking at the best free options the best paid options my ranking for quality of actual models and my ranking for the value of the overall websites I've omitted a lot of websites here that I just don't think are worth talking about at all and of course you know as 2024 kicks in and new AI art websites come out or they get updated these rankings could change and also if you think that I've missed anything in this video please feel free to leave a comment down below oh and keep in mind a lot of this is based on my own opinion my own experience and a lot of times especially when it comes to these AI art image generators well it all depends on what you actually want to generate because a lot of them are really good in some areas and really bad in other areas hopefully whether you're brand new to the AI image generation world or are somewhat of a veteran like myself there will be something new for you to discover in today's video without further Ado I want to start out by diving into what I believe are the best free options ways that you can generate with AI for entirely free let's kick things off with my top free AI image generator now this is actually going to be Dolly 3 made by open AI but in all of the Microsoft configurations it's a little bit confusing but Microsoft has a very close connection with open Ai and well in a lot of Microsoft AI Based Services that are provided by open AI they're essentially just footing the bill allowing you to generate anything you want with doy 3 for entirely free now now there are some limitations in comparison to generating with the paid version of dolly3 which we will get to Microsoft B image Creator this one gives you a certain amount of boosts for the day and then when you run out of boosts your image generation just takes a little bit longer so theoretically we have infinite dolly3 image Generations all we have to do is sign in with a Microsoft account to Bing image creator of course being dolly3 well the quality of your Generations is going to be pretty darn good it understands text quite fantastically the only downside is that we can't change our aspect ratio we have less fine- tuneability and control and it's being hosted by Microsoft so there's some weird little quality discrepancies in your actual image it's compressed what's great is that you can do some relatively uncensored stuff such as famous characters Dolly 3 in the Microsoft formats gives you a lot more diversity in the understanding of your complex text prompts and descriptions the ability to do text on the actual images themselves which is a pretty big deal as well as photo realism whenever one of my friends who's more of a novice in the AI art generation space asks me where they can get started I often recommend Microsoft bang image Creator because it's free and you get lots of really great images whether you're a beginner or a more advanced user like I said there are some other places you can use doly 3 for entirely free through Microsoft where they're just footing the bill another one of those places is Microsoft designer they have automatic background removal generative erase but they do have just the ability to generate really nice images with Dolly 3 and the final Microsoft place where you can use Dolly 3 for entirely free is going to be Microsoft co-pilot or Bing chat this is my least favorable option but it does have access to that dolly3 API you have to talk to the chat bot and it will you know make the prompt for you and everything which can be good for beginners because it might know how to prompt better than a beginner but I think it takes a little bit of the magic away from a AI image generation like learning to prompt AI image generation and I feel like it takes a little bit longer at times and it will often disagree with your prompt more often if you want to make something a little bit more risque or with copyrighted characters but yeah as you can see we generated plenty of really good images just by chatting with the Bing chatbot by no means a bad place to generate I think I just prefer image Creator over the rest all right guys so next up we have got to talk about stable diffusion XL and stable diffusion as a whole because it's a whole another world out there for AI image generation if you want to dive deep into this stuff if you really want to do something specific stable diffusion is often times going to be your best bet I'll try to make this very simple for those of you who are new to image generation stable diffusion is not like the other generators it's open source it's free for anyone to download and manipulate right down to the source code this has allowed people to train their own little stable diffusion models and make all kinds of different stuff there are little stable diffusion luras or tiny models for photo realism or anime style pixel art style it gets really complex and really deep but generating with stable diffusion in your own models can be a little bit more complicated there's various tutorials for that I actually have some tutorials on this channel that I'll link down below but yeah essentially more or less stable diffusion allows you a whole another level of controlability and Nuance than you would get from a normal closed Source AI image generator in this sense you can't really compare something like Dolly 3 or mid journey to stable diffusion because it's just a whole another world but it is a wonderful free option so this is what I'm going to provide to you guys in regards to stable diffusion this is civid ai.com the best place in my opinion to get any of these fancy little stable diffusion trained models you can already see right off the bat we have a model just for Baymax from Big Hero 6 uh models just for Ron Weasley from Harry Potter I mean the list of models goes on and on and on and these can all be applied to stable defusion if you want to really dive head first and dive deep into AI art generation this could be a pretty great place to start but yeah civit AI is entirely free and they have a plethora of models from their amazing Community to download from a little bit more on the complicated side but it absolutely can be worth it with our little stable diffusion bird and bees talk out of the way let's check out some of the websites that combine stable diffusion and other models together or use stable diffusion in their own special way to do AI art image generation in their own way again guys we're still only talking about the free websites here welcome everybody to playground ai.com this is one of the most popular and one of my favorite websites to generate AI art with again you can do AI art image generation for entirely free here on playground there are so many options they have a fantastic interface and they even have their own special model now a lot of these websites such as playground are going to have subscription plans but they do all have at least the ability to create some amount of free images so playground here allows you 500 images per day for entirely free which is pretty darn cool and then you know they've got their other plans which you can check out if you're so inclined but if you're going to get a plan to anything I recommend you try out all these websites for yourself first and then decide upon the one that you like playground is fantastic for beginners they have these little tutorials up in the corner that you can follow to start to generate they've got different methods of generation that are all pretty simple their main way of generating is this canvas tabes things so if I just you know do a quick little generation we can see something pops up and then it moves right next to it to do my next canvas generation but I can actually overlap these if I want and actually combine Generations right off the bat so it's a really nuanced and creative way to actually make AI art and it's very seamless and easy to use like you can see these two generations are combined together here that's a pretty special thing to do and it just does it all seamlessly here on playground AI in 500 a day for entirely free man that is impressive now they do have a more traditional interface which I also love a lot which is their board interface you can see this is just a little bit more complex than generating with Dolly 3 for example because we have a bunch of settings we can tweak the settings are all very user friendly though and you'll notice there's a lot of similar settings across all of these AI image generation websites because a lot of them are just using stable diffusion XL in various models made for stable diffusion XL that's how powerful stable diffusion is in this uh AI Art Space so yeah just a very quick rundown of playground AI first up we have filters here so we can do animated night Vision anime style delicate detail Ultra lighting all of these different styles you can apply obviously your main prompt you can actually expand the prompt which will use AI to improve upon your prompt you already have really cool idea you can actually do a negative prompt here or an exclusion from your image so if there's specific things you don't want in your image you can actually Define them out we can also do image to image which is when we upload an image of you know our self and then do an AI generation off of that if you're new that's a whole another can of worms that can be dived into later I wouldn't Focus too much on image to image at the moment and they also have control traits as well which is another really cool option again this is a little bit more in- depth if you're new I wouldn't worry too much about that on the other side here you can see we can change the model stable diff Fusion XL is the default one just stable the fusion XL without any of those extra added-on models well it can be very powerful I mean this is just raw stable diffusion XL but with those added models you can see things can get changed up pretty easily they actually do have their own special model here on playground which is called playground V2 playground V2 is not a bad model by any means I wouldn't say it's any better than stable diffusion XL though but it's cool that they do at least have it and it is a decent model you can see here are some image Generations from playground V2 with the vibrant glass filter of course you do number of images at once up to four with the free plan here on playground sampler quality and details impr prompt guidance resolution all more advanced options if you're more advanced user so yeah that is playground AI I think it's fantastic pretty great free website to generate AI art just a reminder guys none of these are ranked yet that is coming later in the video right now I'm just talking about First the free ones then the paid ones and then we'll get into ranking next up guys we have idiogram ai this one made a little bit of a splash earlier in the year for being the first mainstream model to do text quite well as you can see there's a lot of text going on with this model yeah this is actually its own separ separate model no stable diffusion and it is pretty darn good does photo realism does all these different characters it's also relatively uncensored which is pretty nice to see with idiogram AI you can do 100 images per day for entirely free or about 25 prompts so you don't really get to create very much for entirely free but I will say that this model does have some clear benefits over you know stable defusion in some small areas like text generation and if you wanted to you know make a logo for your business hey could be a pretty good option not much to say other than it's a pretty cool little model with a decent free plan so next up we have to talk about Leonardo AI this is another phenomenal AI image Generation website they have their own custom models based off of stable diffusion that are pretty popular at this point you still do get 150 daily tokens not as good as playground AI in the Free World but still a phenomenal website nevertheless a huge majority of people do prefer for Leonardo AI because it does a lot of little things very well they have things like prompt magic or Alchemy V2 or photo realism lots of interesting ways to change up your Generations honestly their user interface isn't half bad but I do think that is a little bit more confusing than something like playground AI but that doesn't mean that it's not possible to get a really good generation a little bit quicker sometimes than playground AI because of all those little alchemy or photo real ISM tricks they can add in that you just don't see with something like playground AI again they also have image guidance this is going to be your image to image a little bit more on the complicated side not going to dive right into that either way I would have absolutely zero quarrels with anyone who said something like Hey Leonardo AI is my favorite place to generate AI art I like their options I like what they do better than other companies and I think there's no denying that a lot of the outputs you get from Leonardo are simply mystifyingly good just a little a little bit more limited in terms of how much you can actually do and generate per day for entirely free all right now things are getting interesting this is meta's Imagine This is an entirely free AI image generator made by meta or Facebook and it's actually pretty darn good but it's not competitive in my opinion with something like Dolly 3 or stable diffusion XL still you get some pretty good images for entirely free it's just another place to go ahead and try out AI image generation a lot of these big companies like Facebook are just going to foot the bill you get some interesting stuff out of this one man I think the quality is not too bad with meta imagine but it's not as good as some of the other generators that are out there so yeah I mean this is just worth it to play around and see what you can get like I said the image generation space has grown and evolved a lot there are so many new generators that are worth talking about and entirely free but yeah not too many features other than some basic generation let's move on to the next thing this is very similar to Facebook's generator but it's actually by Google and I think this one's a little bit better in the quality Department to be honest this is Google's imag this one was hyped up for a very long time with a big Google paper and website but was never released and then we actually got access to it and by that time Dolly 3 already existed so you know in my opinion it's not as good as Dolly 3 but it's still again entirely free to generate you can see this is actually pretty good in terms of doing like a copyrighted character like this is not that easy to do even though they look like stuffed animals I'm pretty impressed by the quality that you actually get out of imagin it does a very wide variety of generations all for entirely free it's one of the more competitive image generators that we've seen again not too much in terms of the options department or changing your image it does have this little option to help you write with more detail very similar to the AI prompting we see with Leonardo or playground but other than that you don't really get much in the way of options but I think you guys will end up seeing that this thing does actually rank very high in the actual overall quality ranking because well it is actually quite good so next up guys we're talking about Mage dospace this is kind of an oldfashioned one but it is pretty cool as you can see this is just basically access to your basic stable diffusion you have some basic settings you can change around what's cool is that you have unlimited generations for entirely free no matter what with stable diffusion pretty cool you can also enhance or upscale these Generations on limited for entirely free but if you want access to any specific models or the Mage space exclusive models you'll have to upgrade to one of their plans so for entirely free this is more or less just some basic free stable diffusion I think this is at least an honorable mention especially because with the prop plan it's actually quite good not to mention they do have an explore page where you can kind of see this little social media tab see other people's Generations playground AI also has this and the next generator we're talking about does too so it's nothing new but still nice to see everything sort of categorized in its own special way so next up we're talking about Lexa doart this is more of an AI search page but they actually have their own AI image generator which is not too bad their aperture v3.5 model again this is pretty simple generation overall I don't think this is the best generator in the world but again it is something you can access for entirely free you only can generate 100 images per week though in my opinion Mexico here it's great for searching for AI but in terms of being an actual website to generate AI art it's one of the weaker ones like I said though the search for uh AI art on this it's very easy to find something cool lots of super weird stuff on here that's Lexica at least it's an honorable mention in comparison to the rest not going to spend too much time on it if I had one thing to say about it it's the best AI image Searcher out there moving on to an absolute legend in the AI art generation space crayon this is an air generator that started out as free and has remained free for the whole time they do now have paid plans but you can still generate for free by modern standards it's not really the best but you can still do a lot of fun stuff with it I think this one at least deserves an honorable mention because while it does take like a minute to do a generation they have like this cool Inspirations tab where you can see stuff the community has generated so far and it is its own custom crayon model which is often times trained on a lot of stuff that stable diffusion or Dolly 3 might not be so if you're looking to do something really specific you might find that crayon is oddly suited towards that such as a lot of these cursed meme generations and oh yeah you get nine outputs at once which is pretty darn cool although like I said I mean it's a kind of an outdated model it's not the best wo so that is everything that is free guys that could have been a video all in itself but we're also covering the paid stuff how it compares to the free stuff and then of course the ranking for quality and the websites I'd also like to mention we are not covering IOS and Android apps there's some really great free generators that also exist for IOS and Android but we're just not talking about those today let's dive into the solely paid options first up I think you all know that we have to talk about mid Journey they just launched their brand new V6 which has kind of been taking the AI Art Space by storm there's no free trial of mid Journey it's only paid starting at 10 bucks a month but man I think the generation do speak for themselves mid journey is one of those top competitors I did a whole video about mid Journey V6 it absolutely competes with DOL E3 I mean check out this pixel art that is not easy to do pixel art has classically been very difficult for image generators we can also see that text is now entirely possible with mid Journey V6 as well producing some generations that are simply jaw-dropping like this one blew my mind a perfect Mario and a perfect mid Journey V6 all in like a particular style like really jaw-dropping and of course that classic level of photo realism that we've known since mid Journey V5 that is just out of this world quality I mean truly phenomenal stuff there is no disagreeing that mid journey is one of the best and a lot and I mean a lot of people prefer it over anything else especially this brand new V6 model let's talk about the payment plans because mid Journey like I said is only paid right now no unlimited option the lowest which is 10 bucks a month grants you access to 200 Generations per month which is not a lot at all that is barely anything if you move up to 30 bucks a month you do get 15 hours of fast Generations which is quite a lot more than you get with the basic plan but it's still that's starts at 30 bucks a month and you can get a DE to Leonardo or playground AI plan for 30 bucks a month so we can talk about how mid Journey might compare against Leonardo or playground like there's a lot of features that you get with Leonardo and playground that you just don't get with mid Journey plus not to mention the mid Journey website doesn't allow you to generate yet you still have to generate inside Discord which means you need to make a Discord account and actually generate inside of Discord with a slash imagine command super annoying even though mid journey is extremely expensive they are very promising V6 has turned out so far to be one of the best generators we've ever seen and it's only an alpha with Nick St Pierre on Twitter here noting that in their office hours they stated that a big update to V6 was actually coming that improves upon V6 is ability to generate text even more mid journey is just one of those AI art generators that I end up just keep coming back to because they have some brand new feature that blew me away or most likely their model is doing something that no other model can do just yet like make perfect replications of almost any animation style as proper prompts or points out in this example I mean the quality of this thing is just absolutely off the charts the replication of these different animation Styles is probably enough to get them sued let's be honest all right guys let's talk about Dolly 3 in chat GPT yes I know it's a little bit strange but Dolly 3 is both a free and paid option to get access to Dolly 3 inside of chat GPT you'll need to subscribe to chat GPT plus which is 20 bucks a month and that 20 bucks a month gets you a lot more than just Dolly 3 mind you so it's kind of like this overall bigger AI plan that just happens to include Dolly 3 the main difference here that you can actually just use chat GPT to generate your images for you so make me an image of a cat who is very smart and it's going to take this General depiction I gave it and actually create a prompt around it because it's a language model and it can just do that so this gives us a new way of actually interacting with the model where the llm the chat GPT acts as an intermediary to create better images for us and we can give the llm feedback back and it will remake your prompt for you there's some benefits and downsides to this but you can see it generates a pretty uh descript prompt here and you do get a very good image out of it I will say that there's also the option here of doing 16x9 and 9 by6 images which you cannot do with any of the Microsoft versions of Dolly 3 so if you want to do a 9x6 or a 16x9 aspect ratio Dolly 3 you got to do it inside of chat GPT but man I got to say the images do come out really really nice no complaint from me here I love dolly3 what's been popular on the internet lately has been the ability to do something like this make the cat even smarter and the large language model Chan GPT is going to try to prompt around your feedback taking an already very astute and smart cat and making him somehow even smarter and you can just push things to the Max and get some really wild crazy images it's a fun way to interact and explore with the models however I do think that there is some self-discovery that you can miss by interacting with the model in this way even smarter dang that is one smart cat it's also worth mentioning and noting that I think the best human prompters out there are actually better at prompting than chat GPT could ever hope to be so while this thing is good probably for beginners the more advanced user is going to find it a little bit limiting sometimes my God it's gigabrain cat give him the galactic brain that is the galactic level intellect that I was looking for anyway guys let's go ahead and move on so next up we're talking about Adobe Firefly and this is actually a really good image generation model as you can see you get monthly generative credits we have 250 credits per month technically yeah that is free but I don't know they just had a paid plan and they changed it so they're always changing things Adobe is usually expensive so just to be safe it's here in the paid section and 250 credits for free a month is not much at all anyways this generator actually kind of lends itself towards photo realism and realistic images in general as you can see we did some close-up photos here and they turn out pretty darn good it can do art as well of course don't discount it on that front and the art it generates absolutely is very good but like I said you can actually change things like the color and tone that are very specific or the very specific lighting like I want to see dramatic light for this image and I want the composition to be shallow depth of field I can have photo settings that are very specific so I can have a low or high f- stop I can even change the shutter speed or the field of view there's all kinds of really advanced settings that you get with Adobe Firefly that you just don't see with a lot of other image generators honestly I would have to say that other than mid Journey this is going to be your best generator in terms of making something that is a photograph photo realism I think mid Journey actually has more coherency than Adobe Firefly most of the time but the amount of uh you know specific Crea of control the ability to change the shutter speed the ability to change the aperture that's something you can absolutely try with mid Journey but I don't think it's going to work as well as it works inside of adobe Firefly it's a really solid generator and I think people often Overlook it mostly just because it's not like a nice free place to generate 250 Generations a month is absolutely nothing oh yeah don't forget they also have all of these crazy effects you can combine painting bokeh effect synth wave different materials different Concepts can really get crazy and make some stunning ideas that you just normally wouldn't be able to do with a lot of those other generators so often overlooked maybe for some good reasons Adobe Firefly so that's it for just art generation I want to run through some websites that are honorable mentions they do image generation in a different kind of way they do specific things with it and I want to talk about them these are also all going to be paid but they are absolutely worth talking about and they're a big part of the image generation space so guys let's talk about Korea AI which is currently in open Beta they actually have an entirely free upscale and enhancer which is pretty awesome so the real-time generation with Korea is pretty amazing essentially our output is over here on the right and on this side we can just draw and things are going to change in real time and it's going to you know really closely resemble that original image that you are basing off from as well as Tak in whatever text prompt you want it is amazing stuff so I'm doing a tabby cat on the beach here I'm going to make orange for my tabby cat and then draw him you know right in the center there is our tabby cat oh his head is going to be up here it's going to be pretty big there now we got his big head he's also got a lump but you can kind of see how this goes you can kind of evolve your image over time and just generate some really unique and cool stuff it's a new way of interacting with AI art it doesn't really fit in with the rest of the AI websites that's kind of why it's in its own category but it is really really darn cool and it's so impressive it's absolutely one of the most fun ways to interact with AI art and I just had to at least mention it in today's video by the way you can do something very very similar to this essentially real-time AI art on your own machine at home I have a full tutorial on this that is going to be linked down below local realtime art for entirely free but I'll admit yes Korea is just a little bit faster next up guys we have magnific ai this is entirely paid and the paid plan is ridiculous 40 bucks a month to access magnific AI upscaling it's absolutely absurd but I got to say it is the best AI upscaler I have ever used and ever tried and it produces some pretty insane results you can take your Dolly 3 generation stable diffusion generations and plop them into this upscaler and really enhance them turn them into something you can really share with your friends and actually use in a lot of use cases it might be expensive but I have to admit it is pretty darn good take this t-shirt design for example absolutely crushes that upscale look how detailed it is that's why this deserves at least a mention here upscaling is a huge part of AI image generation especially if you want to get something really nice something that you can look at in full detail cuz a lot of image generation is still low res it might be expensive but man is it ever good hard to beat that detail enhancing magnific next up we have promi another honorable mention this does have some limited free generation per month as well but it's like 200 gens per month nothing crazy the reason it's an honorable mention here is that they have this insane imageo image generation where you can take like a little sketch rendering here and convert it into something that really resembles the original sketch it's pretty darn cool this page is really meant for like interior designing relighting things erasing and replacing objects you can do text effects even create your own AI models it's just an AI site that has taken image to image to a whole another level and I really appreciate what they do even if their paid plans here are a little bit expensive and finally we have another honorable mention here open art this website is pretty good you have a lot of community models that you can actually test out and create with here again this is a lot of like that stable diffusion stuff where it's very specific models created for specific tasks but they're very good at those tasks so it's kind of like this little creation community site where you can actually train your own model and upload it to open art it's only like 12 bucks a month for the minimum pack which actually gives you 5,000 credits a month that's pretty cheap however you don't really get anything for free here it's a one-time trial of 50 credits to experience all of their features which I'm not really a huge fan of but they have gone deep and created this ability where you can upload your own custom model and anyone can use it on this website the only caveat being that it is not entirely free they also have a lot of those fun apps like the QR code maker background remove removal in painting sticker Creator image blending variations it's a good website it's worth mentioning not to be forgotten and by the way guys this is one of the only websites that actually allows you to train your own model whether it be your face your specific character you might have created a specific object or just a style in general those are all of the best free and paid websites that I think are worth mentioning some of their quirks and differences in between each other and now we're going to get into their rank King my opinion on where they fall in between each other the best in terms of quality and then the best in terms of the value that you actually get keep in mind this is my own opinion you might have different opinions and that's okay we can discuss them in the comments respectfully and of course you know as 2024 goes on these might change generators might get updated or new ones might come out for those who might be watching in the future maybe leave a comment about it let's go ahead and dive in all right let's start off by ranking all of the different models I'm going to start with the worst first and move up to my number one again this is my opinion you might have a different opinion we can absolutely have a respectful discussion about it in the comments below all right this is going to be the top 11 models coming in at number 11 is going to be crayon this model is the worst of the bunch today it doesn't do very coherent things at all but it does have a lot of understanding towards pop culture references that a lot of the other AI generators don't crayon is a fun and playful model but it's not really good for more than just just funny memes next up sitting at number 10 is going to be playground's custom model playground V2 this model to me isn't anything really too special very similar to the base stable diffusion XL I think maybe in some areas it will add a little bit of extra Pizzazz or a little bit of extra flare to the images but nothing better in terms of the coherency text following Department from what I can tell it can't capture any more elements from your prompt than stable diffusion could in fact I think that playground V2 is strongly based upon stable diffusion it might be a fine tuning of it moving on to number 9 this is going to be the custom Lexica aperture model V 3.5 now this model I think requires a little bit of prompting to get something good out of but I do think it has more unique capabilities than playground ai's V2 model I've seen a lot of really unique images that come out of it especially made by the community and it also has this ability to do certain Styles really really well which I've noticed Lexica aperture has also been in development a little bit longer than playground ai's models have been it's definitely a fun model to play around with next up at number eight is going to be idiogram AI this is that model that kind of made text generation a little bit mainstream it was very impressive when it launched it does very coherent text a lot of the time impressively and it actually does a pretty good job at combining a lot of the different elements from your prompt into the scene however it lacks in this crystallization Department the Fidelity the actual rendering of the images are often blurry or messy it kind of misses a little bit of that refinement that you'll see with some of the better models on this list this space evolves very rapidly and what used to be like an a tier A+ tier AI is now like a d tier AI big points to this one for remaining relatively uncensored something that was there when this launched and is still there today for the most part all right this next one was very difficult to place this is number seven on our list meta imagine this is a good model for sure it's better than idiogram in terms of that Fidelity Department know it can't do text as well as idiogram AI but in this sense I feel like idiogram AI is kind of a one-trick pony in comparison to a degree all of these AIS are pretty breathtaking after all meta imagine just has this ability to create copyrighted characters it's a little bit more uncensored than a lot of the other AI the infinite generation is also really nice with the The Meta imagine and this model does have the versatility of being integrated into like Instagram for example next up we're moving on to Leonardo ai's custom models these really blew me away they're quite fantastic and they're like stable diffusion with a little bit of flare very similar to like playground AI for example but they've just been working on their custom models a lot longer than playground Ai and I think they are much much better there's a level of fidelity and quality and high resolution that comes with Leonardo custom models that you just don't see from the other stable diffusion based models on this list it might not have the versatility of the top models on this list but it absolutely does produce beautiful gorgeous colorful High Fidelity imagery and there's just something really yummy about that all right number five on our list is going to be adobe's Firefly specifically the new V2 model it's a very very impressive model indeed now we're getting up to some of the big players this model has an uncanny ability to replicate real life images it has a level of fine- tuneability and control ability that you often don't see with these Big Brand company image generators it has decent prompt understanding as well doesn't really do any good text to be honest which is kind of the reason why idiogram is such an outlier today but I still think that adobe Firefly has a lot more usability than idiogram AI for example you can truly make some images that you would be able to use and its creative exploratory aspect is actually really really high up there more so than you see with a lot of other generators that goes a long way number four on our list is going to be Google imag in 2 pretty surprising I know but this model was hyped up for a very long time it's more well known as just Google Imaging but you're using that Google image in 2 API this model is very good it has very good understanding of your prompts it can do some basic text generation as well shockingly knows famous characters very well and it can be very senstive at times but also be very uncensored there's ways that you can trick this thing into actually generating what you want people especially on my Discord server have been having a lot of fun with Google image in too it produces some pretty phenomenal results and can also do photo realism pretty much as well as Adobe Firefly although it doesn't have the same level of fine- tuneability and control I think it's no surprise here that Google is able to produce an AI image generation model that understands the world so well and is so high up on our list today they are the ones who created Google Images after all moving up to number three on our list this is the most difficult AI to place on this list because it just has so many facets for a lot of people this is going to be your number one and I totally understand why this is where I put stable diffusion XL number three it is a very very good model it does minimal text but it understands prompts in a pretty holistic way and being open source it just has so many more capabilities than you see with the rest of the lineup again like I said if this was your number one I totally get why from the fact that people fine-tune this model all the time and make their own custom little versions that do very specific tasks to just the mere fact that this has kind of propelled AI image generation to new heights that we've never seen before it's the most widely used and widely explored model out there and it deserves a top spot on anyone list all right moving on to number two on our list this is a shocking finish because I honestly didn't know if it was going to make it up here but they absolutely did pull it off this is mid Journey's V6 model and right now we only have access to the alpha and it's already up to number two on my list but this model is absolutely phenomenal it does really good text it does really good characters you saw like Flawless replications of different art styles from pop culture it also does insane photo realism again if this was number one on your list I totally understand as well mid journey is fantastic but V6 they really blew it out of the water and I'm really impressed by what they've been able to do one of the most versatile models on our list today and it really holds up as just a single model again I think mid Journey kind of loses a lot of points in comparison to other models because it's just so darn expensive with how much it costs it starts to make it a very tough sell and finally coming in at our number one spot I think you guys already know it's going to be dolly3 yeah I think that Dolly 3 has the best quality out of any of the image generators it has the best text understanding hands down if you use it through the Microsoft methods that we discussed in the beginning of the video you will be able to create pop culture characters fairly regularly famous people is another story I mean you can kind of get around it sometimes certainly when it first released you could and it was pretty insane at that weirdly Dolly 3 does kind of fall behind in photo realism but if you prompt it right you absolutely can get photorealistic results that compete with mid Journey V6 stable diffusion or Google Imaging but Dolly 3 is just so easy to use out of the box it understands text better than any other model I've seen to date and it is just so potent and Powerful somehow they've managed to capture that same Allure of this aesthetically pleasing quality that mid Journey has while keeping like this insane level of just Universal understanding of our world Dolly 3 represents the next generation of image generation It's A Step Above stable diffusion XL mid Journey V6 can compete with it sometimes but Dolly 3 was here first and I still think that it understands prompts better than any other model hands down and that is very very important it can combine like 10 different elements from a prompt into one again some of you might have your own different opinions for where these different AI generators fall one thing that really makes dolly3 stand out to me is the fact that you can just do infinite generations for free through Microsoft I mean that is absolutely crazy that this generator this really really good one is just totally free mid Journey you have to pay at least 10 bucks a month and you don't get many generations at that with Dolly 3 you can just start generating dolly3 is the king of my list it is my favorite generator it is the one that I use the most and I get the most use out of in terms of creating stuff for my Channel or making images for my friends it's my go-to so that's going to be my model rankup again you know share your own rankings down in the comments below I would love to hear your thoughts now I want to go ahead and move on to the value ranking for the websites and the models as a whole so this is going to be a much longer longer list so finally I want to take you guys through my list of perceived value for these AI websites coming in at number 16 is going to be crayon it still is useful for creating fun memes and doing pop culture stuff but it's kind of Blobby the model isn't that great it's still worth mentioning and it's still worth knowing about you're just not going to get that much out of it in general next next up is going to be open art this is a website that I just haven't used in a long time and it does have some special capabilities that help it stand out but they're very very Niche I think if you want to create your own custom model with your own face there are other options than open art but open art does have that feature where you can share them with each other and that's just something you don't really see so open art that's going to be my number 15 spot here it's a really good website lots of good features but I think it's just a little bit too Niche to be high up on this list next up is going to be magnific com com in at number 14 this website is very good it's the best AI upscaling I've ever seen but again we're looking at this from a value perspective they charge you $40 a month for the minimum magnific plan that is just out of this world insane that is horrible I don't know what they were thinking when they're charging for that but that is not worth it at all so while it is really really good AI upscaling cre AI does upscaling that's very similar for entirely free so magnific is useful and I think when you're making AI art you want to learn to upscale but for 40 bucks a month it's just getting put right down at the bottom of this list today number 13 here on our list is going to be Adobe Firefly this is kind of the same reason they're only giving you a limited amount of generations per month right now I swear they had a payment plan but for some reason I could not find it on their website so it seems like they don't really know what to do in terms of a payment system for Adobe Firefly classically Adobe products are extremely expensive so this one is just towards the bottom of the list because well it really only does photo realism that well and does have nice controls for photo realism and a nice interface there's plenty of other AIS that are also great at photo realism and I just don't think it has a lot of value over those other AI generators for what you're getting through the Adobe Firefly website next up is going to be meta imagine at number 12 on our list imagine is entirely free which is really nice but there's like no features on the website sure you can do stickers in Instagram which is pretty cool but I just think that because it's free it gets a little bump up from Adobe Firefly other than that it's an okay model pretty smack dab in the center of my previous list if it had a better website with more fine-tunability and control maybe it would be higher on the list but for now it's just kind of a little Tech demo where you can generate images with meta imagine Google imagin 2 is going to be number 11 on our list while it is a very good model and it almost tops the charts of our previous list image in 2 is just not a great website to work with again it's kind of similar to meta imagin where you just have a prompt box it's it's kind of like a little Tech demo there's not a lot of fine- tuneability and control with your prompts they do have one feature above meta imagine which is that AI prompting which is pretty cool but other than that it's just like generate okay got some images generate okay got some images the model Google image into is pretty standout good but I just think for usability there's so many websites that actually offer a lot of different structure and changes that you can make to your images that give you so much fine-tune control next up number 10 on our list is going to be prome a AI this website is phenomenal it allows a new level of image to image control that I've never seen before so many different types of image to image their plans are a little expensive but again they're doing things that I just haven't seen before with image to image taking a sketch and converting it into a real image doing fully custom reusable AI models Lexica aperture is Right Above It at number nine Lexica aperture is only here because it is straight up the best AI image search that you can get it's very quick and you can compare images and get really creative ideas for AI image generation from Lexica very quickly you can also see what a lot of the stable diffusion or lexic model Generations can look like and you can build off of them easily it is a massive repository of AI images as well so there's just so so much on that website to explore and it's very very useful for actually using in your AI creation workflow and just one tick above that in the number eight spot is going to be Mage dospace this website's pretty oldfashioned it has to sit at number eight just because it's offering infinite stable diffusion image generation not a lot of websites offer that it is completely infinite completely free and honestly for a very reasonable monthly sum you can get access to 160 plus custom models that's pretty darn good good value with Mage do space for sure in my eyes it's kind of like that classic stable diffusion image generation but now you get to access a ton of models for a reasonable price number seven on my list is going to be Korea AI shockingly I actually did put this above Lexica Mage space prome Korea does something that most of these other websites just don't do and that is real-time image generation an entirely free AI upscaling like I was saying in comparison to magnific the upscaling is not as good but it's totally free and upscaling is a huge part of AI image generation their plan for Kaa does allow infinite generation but it is kind of expensive 30 bucks a month but that real-time generation is just so slick and so cool there's a lot of value to be had through Korea AI at this moment this could shift in the future but for now Korea gets a nice spot on my list next up on my list is going to be idiogram AI at number six I know a lot of you might be shocked idiogram AI isn't that good a lot of the time but it is the second cheapest way you can do text generation with AI and I think that's pretty huge that gets it a pretty high position on my list personally if you wanted to do logo design for example completely free with idiogram AI and of course it has that relatively uncensored level of Pop Culture references next up on our list is going to be mid Journey mid Journey has a lot of capabilities it's a really nice AI tool in the new V6 model as I just discussed is extremely versatile and quite amazing however mid journey is super expensive giving you only 200 Generations a month for 10 bucks a month pretty bad deal and the next plan is 30 bucks a month which is quite quite a lot I think mid Journey V6 is super useful it's just a little bit too expensive and that's why it's number five on my list today if mid Journey had some intermediary plan that was let's say 15 or 20 bucks a month maybe it would get a little bit higher up on the list there's a lot of value to be had there and they do produce very promising models coming in at number four is Leonardo AI an absolute King in the AI Generation website World they have all kinds of different custom models the custom Leonardo model stable diffusion models lots of little tricks to enhance your images Leonardo pretty great but I don't think that they give as much for free as other websites might that's kind of what's keeping it from being even higher up on the list but hey number four is still pretty high overall Leonardo AI has really impressed me they're able to do things that other websites aren't able to do with AI image generation they also have real-time generation really good image to image Leonardo is packed full of features and at a decent price as well for their different plans again what's holding them back especially in comparison to number three is just the amount that you get for entirely free coming in at number three is playground AI a really phenomenal website to generate art 500 images a day for free is incredible access to a ton of different custom models as well as that custom playground model which you know it's not a showstopper but it's there also have a really nice ability to share and explore other people's Generations which is pretty cool and byy far in my opinion playground AI has the best user interface for generating you can do it either on a board or you can do that canvas thing playground AI also has this ability to do joint Creations where multiple people can be generating at once on one single board which is really amazing so I think that playground AI has really high value you get a lot for entirely free their plans are also a really really stinking good deal while they might not have some of the more creative AI generation ideas that Leonardo has they still have quite a few very very competitive in number three on my list coming in at number two on my value list is of course going to be Dolly 3 but not in chat GPT because that is 20 bucks a month if you just want an image generation I don't recommend you go buy a chat GPT subscription but inside of the Bing create you know Microsoft designer Bing chat all those Microsoft options for Dolly 3 it's entirely free and it's infinite and it's the best image generation in my opinion so it has to just take a number two spot cuz it's entirely free there's not a lot of options for generating sure but I think that the dolly 3 Model is so darn good that it can kind of override a lot of the creative stuff that you might see with playground AI or Leonardo even the realtime generation there's just so much that you can do with Dolly 3 it's such a versatile model it's able to propel it just by being free to number two on my spot and finally the highest value that you can possibly get in the AI image generation space is in my opinion going to be stable diffusion it's available on a variety of websites for entirely free you can generate at home for entirely free again I have tutorials linked down in the description below for that stable diffusion and all the custom models that are built off of it is just this Titan of the AI art generation space that gives you an insane level of value almost everything that Dolly 3 Google imagin Etc can do really well can be done very well by a custom stable diffusion model so you're just getting so much bang for the buck essentially for free you can harness the power of the open source Community I love stable diffusion sure in comparison to Dolly 3 you're not going to get the same level of coherency and text understanding but you're getting at least you know 60 70% of it again this is my own opinion my own experiences if you guys have something different we can totally talk about it in the comments below but guys that is going to do it for today's video I know it was super long and it's a doozy but I think this is going to be a great introduction for a lot of people a great resource for a lot of people into the AI art generation space I stand by every website that I mentioned in today's video and I think these are all the most relevant ones going into 2024 thank you so much for watching everybody I am at vidpro aai if this video was useful please consider subscribing joining my Discord following me on Twitter Etc we talk about a lot more than just AI image generation on this channel so definitely worth a little peek if this website was helpful to you thanks so much for watching I'll see you in the next one goodbye

Channel: MattVidPro AI

Views: 73,818

Keywords: ai tools, ai art, ai art generator, ai art website, AI Image Generation, Bing Image Creator, Microsoft Designer, Stable Diffusion, Civit AI, SDXL, Playground AI, Ideogram AI, Leonardo AI, Meta Imagine, Google Imagen, Lexica AI, CrAIyon, Midjourney, ChatGPT, Adobe Firefly, Krea AI, Magnific Upscaling, AI Art Models, AI Art Sites, AI Tools Guide, Free AI Tools, AI Image Creation, Image Generation Software, AI Art Generation, AI Image Tools, AI Art Platforms, dall e 3

Id: z_aKLP5pBhY

Length: 51min 22sec (3082 seconds)

Published: Mon Jan 01 2024