hi everyone I want to introduce you to an awesome AI video generator called Minx by a company called Hyo now they've just released a brand new feature that a lot of people have been waiting for and that is image to video so now you can upload your own images and bring them to life and in my opinion Minimax actually has some of the best prompt adherence especially if you're doing text to video but image to video is just as impressive so without further Ado let's get started so from the date of publishing this video you
can still use Minimax completely for free and you can generate unlimited videos however they are actually going to introduce their subscription pricing plan tomorrow so if you guys are watching this video right now go and generate some videos however there is still going to be a free tier so this is what I saw on X you can see with their subscription uh pricing you'll have a free tier and you essentially get a 100 credits for logging in Daily and that allows you to generate three videos because each video uses 30 credits now they also mentioned
over here that new users will receive 3 days of unlimited generations for a limited time so that's quite new if you jumping on board quite early with the stuff you'll still be able to generate unlimited Generations until they decide to turn that off anyway there is going to be a $10 plan over here as well which uses 1,000 credits and then there's a very pricey plan uh which is maybe for people that are enthusiasts or filmmakers and that is at $95 and you get the unlimited plan but I did want to mention that and be
transparent so you guys can see that there is going to to be some sort of pricing involved over here so to access this website just go to this URL when you're on the website just click on sign in and you can sign in with Google so once you've logged in on the homepage you're going to see all of these amazing videos that have actually been generated by Minimax so if I click on on any of these videos I can see the prompt that was used so this is a great way to get some inspiration but
I can also get a general idea of how they structure their prompts to create videos like this right so that definitely gives you some good Insight uh some of these uh these prompts over here you'll notice are in Chinese like this one for example but if you select all of this text you right click and you just select translate selection to English you'll be able to see what that prompt is in English so definely go ahead and look at some of these examples so you can see what Minimax is capable of doing with text to
video Minimax definitely has some of the best prompt adherence especially with text to video you can see over here with this video that I've generated I said that it's a hybrid creature that is half llama and half peacock it walks through a mountainous range and everything I've mentioned over here it's doing exactly that in the video and that's incredible so depending on how detailed you want your prompts to be uh you can really direct and control a lot of the aspects of the video that's going to be generated so let's try text video so here
at the top we have the prompt box now you can type in whatever you want I'm going to type in a mythical hybrid creature that is half Fox and half peacock sitting in a forest and before we click on create I just want to explain something else so you'll see an icon over here called The Prompt Optimizer now you can see when I hover over this it's explaining exactly what it does but just to explain it quickly with this disabled by the way you just left click on here to enable or disable it but with
it disabled uh it's essentially going to adhere to this prompt and it's not going to enhance it or make it more detailed so whatever is typed out here that's exactly what it's going to to try and generate but if I enable this it's going to take this prompt and make it a lot more detailed so it might add camera movement lighting or even an additional setting onto this prompt and this is explained right over here on the Discord so you can pause the video right now if you want to so you can understand exactly what
the prompt Optimizer is doing so I tend to actually leave that off because I wanted to follow my prompt as close as possible so once you're ready just click on create let me make sure this is turned off so yeah just click on create and it's going to generate a video for you so you'll see that this starts spinning and now you'll have access to a section over here called mine so this will be where all of your videos are contained that you are currently generating so as of right now and this is probably going
to change with a paid plan you can only generate one video at a time so you have to wait until this reaches 100% And your video is generated before you can generate another one and the videos that get generated with text to video are going to be in a resolution of 1280x 720 also keep in mind that this AI video generator model is not completely uncensored so don't try and generate videos that include Gore or nudity it's probably going to get blocked so just keep that in mind videos that get generated are 6 seconds long
and there's no way to extend them right now but I'm sure that's going to be introduced at some point so the video has generated and this does bring me to another Point even though Minimax has good prompt adherence it's not perfect right and that's the nature of AI video we have to roll prompts sometimes to get the desired result in this case obviously that's not a hybrid creature but I did do another video like this with the exact same prompt and I did end up getting that result so over here I do have a fox
and a peacock hybrid so you are definitely going to have to reroll some prompts to get the desired result but when it does adhere to The Prompt it does a very good job and now for probably one of the best features image to video so this officially Sol character consistency or just consistency in general because previously you'd have to actually prompt that you'd have to prompt a specific style or characteristics to get something coherent now you can just upload an image and you can generate these amazing videos so to use image to video it's very
simple just find an image that you want to use you can drag and drop it over here directly onto Minimax and then it will start uploading it or you can click on this icon and then click on uh click to upload an image and then find whatever that that image is on your desktop but another cool thing is you can see every image that I've uploaded has basically been saved at the bottom so this makes it very easy if I want to reuse a previous image maybe with a different prompt so my image has been
uploaded and in this case there's two ways to bring this to life so right now I don't even need to type in a prompt I can just click on Create and I can give AI complete freedom to analyze this image and bring it to life or I can be very specific with a prompt and describe to the AI what I actually want to happen with this image in this case I'm just going to type in she starts smiling very simple prompt and then I'm going to click on create so again you just wait for this
to load and like I mentioned previously it's going to generate one video and you have to wait for that video to generate completely before you can do another one so another awesome thing about image to video besides consistency is that we are not restricted to 1280 x 720 as our aspect ratio I've generated videos over here you using image to video that are vertical all right so this could be something for I don't know I don't know like maybe Tik Tok or Instagram story you can see that we can use different aspect ratios which I
think is amazing so the video is generated and like I said the prompt was simple it says she starts smiling and I mean look at that she looks very expressive maybe there's some minor minor issues with her eyes especially towards the end maybe some minor ghosting but I mean this looks extremely expressive this image has officially been brought to life and this is where AI to me personally is just mind-blowing that it can even do stuff like this and this is also way Minimax excels it's with emotion so you can prompt all sorts of emotion
to get different results so for instance with this video The Mona has been naughty right she's been arrested and you can see the prompt over here said she's laughing and then she starts crying however this emotion does look like she's simultaneously laughing and crying I I thought it was quite funny but what really surprised me was the text over here the fact that it's still remaining consistent while it's bending and moving that is quite crazy and then you can see you can make them smile so try all sorts of different emotions smiling scared shocked and
see what you come up with uh this one was also quite funny I did promp she sticks her tongue out but it does look like she's licking in an imaginary ice cream for some reason so this is where you'd have to adjust this prompt maybe say she sticks her tongue out and keeps her tongue out uh and yeah that's all you know uh aim of using AI to AI video generators is that it involves iteration until you get a result that you are happy with Minx 100% excels at anim animals in this case this is
a hybrid monkey and lizard my prompt was just monkey and I mean look at this how it brings it to life that's crazy it's like maybe something from National Geographic uh so definitely give this a go with some animals uh the there's some stuff over here where I've used text to video for instance this horse and peacock hybrid and the fact that it understands the walking motion of a horse is crazy to me now of course there is some inconsistencies in the flowers over here but the way that it's animating this animal to me personally
looks insane so definitely try using Minimax to animate some animals so something also wanted to try with Minimax that actually really surprised me was using an image that stylized and animate it and over here you can see it still retains the consistency of the image while animating it and this is a major plus because a lot of other AI video generators tend to morph into different styles or they make the characters look extremely realistic but over here the style is remaining consistent and this is a very good example this is supposed to look like a
painting but you can see it's still animating it and the style is consistent now these brush Strokes are I mean they're not moving with the rest of the face so it's not 100% perfect but but this is a fantastic result in my opinion that it can even do something like this and here's another example that's imp pop out you can see it's animating the lips there's a little bit of eye movement over there but the style is consistent so give this a try guys try some styliz image images remember sometimes you might not get the
best result on your first try so reroll your prompt until you get something you like so you can even try to animate maybe some movie posters over here I said a woman screaming in fear crying the text scream remain static now the text that is actually included on this poster did remain static but it also disappeared which I think looks really cool but I mean look at the emotion in this video that is insane uh she really looks like she is very terrified if you guys are trying to do these fashion catwalk videos that you
see all over the Internet yes minax can do it it actually understands the walkin cycle quite well and this is the prompt so as long as you say walking down a fashion runway you're going to get some really cool looking videos you can also try using Minx for e-commerce if you're trying to promote any fashion products or maybe you just created some AI fashion and you want to bring it to life uh minax does a really good job and in this case this really surprised me because you can see the person actually turns to their
side and everything remains consistent from the glasses to a hairstyle to the Garment and that is some insane consistency uh so yeah give this a try with fashion as well you can certainly use Minx to promote products as well so I tried this with a perfume bottle but you definitely want to include some sort of camera movement in your prompt if you want it to be more Dynamic yes you should definitely try this with scenery you can get some interesting results but I have noticed with quite a few of my Generations I've got static videos
uh so like the background elements will be animated which looks cool uh but you should definitely include camera movement in the prompt if you want something to look a little bit more Dynamic of course it is not perfect sometimes it actually does not adhere to the prompt and you're going to have to reroll until you get something that you like yes you can also prompt for people to eat food in this case I said the Mona is just eating ice cream and you see that small detail left behind right over here there's actually ice cream
that's left on her lip and that to me actually blew my mind when I saw that but this is still far from perfect because the ice cream itself refills itself so she's got this magical endless ice cream that cannot be finished I did tried with spaghetti and uh well this just looks kind of strange so to be completely honest I do think cing does videos like this better especially where they eating food and over here it seems like you have to have the food in the actual image itself for the prompting to work also don't
forget that you can just upload an image and click on generate without a prompt and see what the AI actually ends up doing uh in this case it analyzed this image it identified a cat it identified a human and and it made them interact with each other and the end result actually looks pretty good I mean there's obviously some inconsistencies in the hands but hands is really difficult for AI especially with AI video generation but the fact that you can do this without even putting in a prompt uh is amazing now in the world of
AI yes hands are still a problem and if we talking about AI video generators hands are well they can be a major problem there can be a lot of inconsistencies so what I've noticed is that this does a really good job with close-ups or torsos or where the hands are not visible although sometimes you know it does the hands pretty well I mean over here there's no issues with the hands uh but in this case I did try you know doing something with this video where she's taking a selfie and these hands are basically possessed
and that that is a really bad result but I tried you know rerolling that prompt and I got something else that was much better so that we've got this real life Mona Lisa taking a selfie but keep in mind that hands well that can still be an issue so in my opinion the best results you get is if the face is actually closer in the image if the face is a bit too far away uh that's where I've noticed the most inconsistent results even with retaining character consistency so just keep that in mind if you've
got shots where you know it's a lot closer to the camera uh chances are you're going to get much better results so the community is already generating some really interesting results and this was my introduction to image to video so this is by Brent Lynch he took one of my photos of my realistic Mona Lisa and he brought it to life so here's the prompt that he used as well so this lady is ecstatic to read the to meet the real Mona Lisa and you can see he's got some other results over here this one
was quite funny she just starts poking her face uh but you can see how this essentially just brings this image completely to Life Next Up we've got Al Sinner and I hope I'm saying that correctly uh but yeah very Dynamic motion over here very cinematic this reminds me of Uncharted uh the video game uh so really cool stuff going on here besides his legs disappearing but I'm sure with maybe another reroll it could be a little bit more consistent but very cool shot and this one over here we've got this vehicle that blows up boom
so cool stuff and he does mention that we should try and push its limits because it can do more than we expect now we've got mria mria Ai and like I've mentioned in the tutorial it animates animals very well and this is just super cute there's another result over here by Cole tasz and a nice shot over here of this tribal man deep in thought and that looks very very realistic I mean even the feathers on his head are gently blowing in the wind and that's just crazy that AI can do that over here we
have suu and they've basically just uploaded an image with no prompt I did translate this so this is an image with no prompt I mean look at this this looks 100% believable right that's extremely realistic and then uh last over here we've got mad pencil and he's done something cool so we've got this head that explodes and boom it looks like it turns into some type of liquid although it would have been cool if maybe the chair and the table actually got drenched in water as well but still interesting result result and then this one
reminds me of Mary Poppins got this granny and this magical forest levitating so yeah the community is doing really cool stuff okay so that's going to be the end of this tutorial I've given you guys a full breakdown of how to generate videos with text and even image to video I've given you some use cases you can see just how awesome this is so try generating videos with simple prompts or complicated prompts and see what you end up getting I really really think uh Minimax has a lot of potential in the AI video generator scene
uh so definitely go and use this right now while it's free if you're watching this video at this present moment give it a try I think it's amazing and yeah create some magic as always thank you so much for watching this video you guys are super awesome stay tuned for some more videos and tutorials and goodbye [Music] [Music]
