 Hello everybody. Welcome back to another AI video. In this one I want to do a comparison video between the just launched SDXL 1.0 model from Stability AI and this is a free open source model versus mid-journey and the way to do this I believe to do it the most fairly is to first off we're going to create a random prompt generator which I'll show you how to do in chat gpt and then once that's done we're going to run the prompts in both SDXL and in mid-journey we're going to compare the images and then you can see what the free version versus the paid version and you'll see how good the free stuff is getting not going to lie SDXL is very very strong and you'll see just how far open source is getting so anyways let's get into this let's create the random prompt generator first all right so the first step is we're in gpt4 this works just as fine in gpt 3.5 I just have four so if I'm paying for it why not use it here is the prompt I want you to act as a random prompt generator for an AI image generator for your prompts please use please use different styles techniques and levels of complexity please create five random prompts and I'm doing this randomly because that way I can't cherry pick what works good and one and not good on the other so as random as possible I'm going to hit enter I'm going to let this run when it comes back I'll start inputting them into the two different systems and then we'll get going all right welcome back so here are the five prompts we got and they're kind of cool a surreal a surreal landscape at twilight where the sky is composed of a vibrant nebula and the trees are made of swirling galaxies wow cool an 18th century bustling market scene with colorful stalls selling exotic fruits various spices and intricate silk fabrics this is as random as it gets fantastic they got a children's book illustration style a cyborg cheetah okay we're getting into some stuff now so let me show you how to input them into sdxl for free and then I'll show you how to do it in mid-journey here we go all right so we'll start with stable diffusion xl because that just launched to do that go to clip drop dot co forward slash stable diffusion as you'll see here I'm logged in of course I will put a link in the description below now let's start entering in the different prompts we're going to head back into chat gpt I'm going to grab the first one I'm just going to copy and paste that in here oops that's not the one we're going to copy and paste it here and then I'm going to hit generate and now when I click that you'll see that you can subscribe for faster generation but you still get it it just takes a few seconds longer and then while that loads I'm going to go into my discord into my mid-journey bot forward slash imagine copying and pasting it in directly the same I'll come back when they're both done we'll do a quick comparison all right welcome back so here is the first prompt run now let's just take a look at the four images it gave us with sdxl this one looks pretty darn good looks good nice swirling nebula the trees the nice it's a little bit uh a little uh less contrasty than the mid-journey one which i'm going to show you right now let's take a look here are the four mid-journey candidates this has a little bit more of a painterly feel but overall they're both very very good I mean obviously some would work better in different situations but I would give it a tie although I'm kind of leaning towards mid-journey with the brighter higher contrast but again you guys decide let's run the next prompt all right let's run the next prompt you'll notice that this time I clicked on the style button and I want to remind you that in stable diffusion here at least on clip drop you can try different styles so there's anime photographic digital art comic book so keep that in mind I'm going to go with the default default style but you can go ahead and change that up if you see fit for example fantasy art would have been pretty good for the last one anyways let's go then 18th century bustling market scene with colorful stalls selling exotic fruits various spices and intricate fabrics or something like this I'm a cat I'm going to hit skip you'll notice there's only 120 images waiting to be processed so right now in fact um clip drop and sdxl is running faster than my mid-journey which I paid for I'm going to go on the record and say that the mid-journey is taking longer all right let's go back eight images waiting to be processed wow that was pretty quick we've even got our first couple okay here we go all right so I see sort of like a 17th century painting if you ask me let's take a look here here's the first one here's the second one okay all right a little bit of tinteretto or if uh I guess if one of the early renaissance painters where we're making them they're all very similar in style and composition if I'm being blunt here let's take a look over here at the mid-journey version in fact we're at 46 percent so I'm gonna have to come back when it's done and show you what they came up with all right here we go so here's what we got in discord and this I mean it has more contrast it's a little bit more vibrance and you know I guess it's a little bit more I mean stylistically it's a little bit I wouldn't say more accurate but there's less it's less abstract but again they're both very very good I'm gonna go back to clip uh I'm gonna go back to clip drop here and you'll notice that this is the same stuff here but I ran this with a different filter here and this is fantastic let me show you what I did here so I'm gonna click on back and this is the same exact same prompt but instead of just clicking it as is I selected photographic now you could put the same you could type in forward slash you know or you can add dash dash photographic or whatever or just put in photographic into mid-journey pardon me but again sdxl even though it doesn't always get it as far maybe as you'd like it for compared to mid-journey you can make a few changes like this one here and you can get some pretty damn good results let's click on the third one all right so for the third one I'm gonna actually start in mid-journey because mid-journey is taking longer than sdxl the free version on clip drop so here we go a hyper realistic portrait of a cyborg cheetah sprinting through a futuristic cityscape to bathe the neon lights how about that that is cool all right let's go back to google chrome here and I'm going to take that exact same prompt I'm going to put it in here and then for stylistic I'm going to leave it at no style but always remember you can go ahead and adjust it also you can adjust the aspect ratio I have a 69 aspect ratio in mid-journey so let's run the same aspect ratio here in stable in sdxl why not it seems fair okay they'll take a few seconds for this to render I'll come back when it's done all right welcome back so look at that now this is awesome like this is like some cyberpunk super cool cheetah I love these yeah these are fantastic images the background you'll notice has a little bit of a blur to it so there's a depth of field applied here and I mean these are just wonderful look at that let's go take a look at discord and see what we got in mid-journey all right these are awesome too completely I mean different to some degree but awesome they've I mean okay guys you can't go wrong with either one of them I'm going to run the last two put them in at the end of the video but you get the gist of it both of them are awesome a little bit in different style clip drop and using sdxl 1.0 is free this mid-journey is obviously a paid for version um yeah you guys decide let me know in the comments what you think thanks for watching