Overview of Free AI Video Tools

There's a lot of amazing free AI video tools out right now, some that are brand new. You can make a character move any way you want, inpaint and upscale, create entire music videos from a prompt. There's of course text to video and image to video. You can create consistent characters, add lip syncing and sound design, all completely free. I want to start with the one I've been having the most fun with, Vigl. You can replace any character with a single input image, and it's super easy. There's endless memeable use cases, a lot have gone viral already, but there's more practical ways to use it too. So let's jump in. I'm in Discord right now, you go to any of these animate channels, and there's two main commands to use this, either with their preset videos or inputting your own. I'll start with the presets, for that we use forward slash animate. Then I'll upload a picture of myself. Then if you use this link right here it will take you to their site with all sorts of preset videos to choose from. Tons of dance moves. You got the Joker. Here's the Lil Yachty clip. I'll try this little dance out. Copy that, paste it in here, then select white or green if you want to remove the background or template to keep the one that's already there and use fine tune on, then submit. Depending on the time of day these may come back in 30 seconds or a few minutes. Then it will look something like Really solid and it's so easy to do. But I want to use my own video instead of a preset. For this, we'll use forward slash mix. Upload that same image. Now I can upload my own video. So this is a short clip from Dance of the Sugar Plum Fairy. I want to use that because, you know, I never took the time to learn ballet. I'll do green so the background is easy to remove and fine tune on. This is what came back which is amazing considering how complex some of the movement is. The model is able to guess all the angles that aren't shown in the input image. Same with if you just use an image of your upper body, it will guess on the legs and feet too. It's not perfect yet, but pretty incredible. Now I need to bring this into a video editor. I'm using Premiere, this could all be done in a free editor like DaVinci Resolve, or even a simpler one like CapCut. I've got the original footage on the timeline. Now I'll drag the green screen footage from Vigl over the top. Then go to Effects and use Ultra Key. Drag that onto the clip to apply it. Then up to the Effect Control Panel, click on the eyedropper next to key color, then click green anywhere on the clip and it's gone. But you can still see the original dancer underneath. To fix that we'll go to runway ml and use the inpainting tool. There's a ton of other tools in here but this is all we need for this. Upload the video and I'll drag that on. Now I'll paint over the character I want removed. Now it thinks for a second then fills in the background. And just from that one keyframe, it will guess what it's supposed to do across the whole video. This shot isn't too complex, so it does it really well. And I can export this. Under the free plan, you can only export up to 720p. That's what this clip was already. But also we're going to upscale this later anyways. Now I can bring that into Premiere and replace the original clip. And it looks great. And a quick tip in Premiere, since the edges are a little hard and don't blend well, Back on the effect control panel, you can click matte cleanup and add a little bit of softness to smooth it out. That helps sell it a little more. So I'll export that. Then I want to upscale to increase the resolution. The best and easiest free upscaler I've found is right in CapCut. It's under magic tools, then video upscaler. Just upload. Upscale 2x takes just a little bit and now we have our high definition clip all finished. Let's check it out. Alright that's amazing. I think I'd make a beautiful ballet dancer. And I know everyone out there is probably asking, how would that look if you were wearing a leotard? Well, here you go. Just beautiful. That's probably too much of a tangent to show that full process here, but for anyone wondering, I uploaded myself as a character reference in Midjourney, added a style reference I found, then face swapped that image using Insight Face Swap. Some of the outtakes trying to get that image were Pretty horrifying. Not sure if I should even show some of them on YouTube. Alright, Vigl is a ton of fun, but I'll show some other limitations to save you some time. Adding a tutu looked funny but has no real like bouncing physics in the fabric. Another area that's difficult is fast moving scenes like this one. It's really difficult for Runway to inpaint it with a constantly changing background like that. There's other solutions but not easy or free. Also if there's something that covers up part of your body in the video, like right here where the chain covers the arm and the wrecking ball covers part of the leg. That would involve some kind of complex masking to fix. Plus this had the same issues with the moving background. There's endless other ways you could use Vigl. It works with all sorts of images. They could be cartoons, Pixar style, or here's an example of one with an illustration style generated with Firefly. I could go on and on messing around with Vigl. It's a ton of fun, but we've got to move on. The fun doesn't stop there. If you want to learn more about taking advantage of AI for your YouTube channel or business, I highly suggest you check out this ebook, Using Generative AI to Scale Your Content Operations. I've put a link in the description for you to download it for free. This is a super comprehensive guide that covers different types of AI tools, how to write prompts for them, and many ways to use AI like to simplify your research, and Breakthrough Writer's Block, which I use it for a lot. I really like that there are specific prompts written out, and I found it particularly helpful that they added some things to watch out for. like bias, risks of plagiarism, and they clarify how Google handles AI content. It's always good to be aware of the limitations and pitfalls as well. This ebook was made by HubSpot, which is today's video sponsor. Big shout out to them for this free resource. Next I want to show another really new free tool called Noisy. You can input the URL of a song, then add a text prompt, and optionally add an image reference, then it will create an entire animated music video. It's still in Discord, but they're launching the web version any day now. So go check to see if it's available there first, depending on when you're watching this video. I've got a song I generated in Udio, which is another free tool. My last video was all about it. This is a metal song about gargoyles. Let's go in for a Dio vibe. I'm going to use this as a reference image. I'll need a URL for it. This process will be much easier when the site launches, but this is the process right now. So I go down to this channel and send it here, then copy the link and save that in a notepad. Now here's my song in Udio. I copy that URL. This also works with Suno, Soundcloud, YouTube, or you could send any MP3 in that same channel and use that URL. Now I'll use forward slash imagine. I'll paste the Udio URL. Now I'll copy that image URL I saved and paste it here. Then add my prompt. Gargoyles atop cathedrals flying through the night. Hit send and wait a couple minutes for it to generate. Once it's done you can click the edit button and it will take you to the site where you can regenerate any clip. I usually just like to generate a couple full versions of the song, then mix and match the clips. Let's see what we got. That is awesome. It's pretty crazy how easy it is to get that whole video. I made a bunch of other styles with Noisy in my last video, which was that deep dive into Udio. Here's a couple clips from that to give a feel of the different styles it can do. I'll just play little bits of each though. There is a house way down in New Orleans I'll call you sweetheart again, even if you drove your car right off that overpass. Now I'll jump into Pixverse. It does text-to-video, image-to-video, and also can do consistent characters. They do have a web app for a better UI, but they have limited generations per day there. Right now you can still generate as many as you want for free in Discord. At least, I haven't hit a limit. So I'll start with text-to-video, which is forward slash create. I'll do... A waterfall flowing into a pool surrounded by trees. Then there's an option to select style. I'll use realistic. Then the aspect ratio 16 9. I'll actually copy the prompt so I can run it in each style for a comparison. And submit. It looks like anime was done first. It comes back with four options. I'll click through each. Those all turned out really well. Then I can either upscale my favorite or run a variation of it to get more similar options to choose from. Let's go check out Realistic. These are good... Not great. Now 3D animation. Those all look great too. The only style it really struggled with was realistic. Different styles work better depending on the prompt. Let's move on to Image to Video. Forward slash, Animate. I've got this nuclear explosion I generated in Midjourney, and I'll drop it in here. Then you can select how much motion you want. It defaults to 5. I'll just leave it there. You can add a prompt. Explosion. Then I'll switch to HD quality, which makes it take a little longer. So while that's working, I'll get another one started. The first one is done. And that looks awesome. So I can click right here to upscale it. Let's check out the other one. That looks great too. Alright, now I'll jump onto the site to show the character feature. You could do this in Discord too. If you click create, you can do all the stuff I just showed. There's even some extra controls for camera movement. But we'll jump over to the character tab. So click create character. Then I'll upload an image of myself as a Pixar character and name it... Kevin. Now when I click create, then click character, I have the option to choose from any character I've uploaded, which is only this one right now. Then I just type a prompt and submit. I'll try a few different ones. So here's all of them next to each other. You can see it's a very consistent character across all the shots. There's another free text-to-video and image-to-video tool called Hyper. I want to show that really quick. It's in Discord and uses the same commands as Pixverse. The downside is the generations are only two seconds. So I'll jump in here and use forward slash create, then run the same prompts. Nuclear explosion, a waterfall flowing into a pool surrounded by trees. Then I'll use forward slash animate and bring in this same image. Let's check these out. So Hyper also does a really good job, but the two seconds is a pretty big limitation. Although if you get creative you can still come up with some cool stuff. I saw this one on Twitter today. Next up I want to show how you can add lip syncing to any of your videos. and also generate sound effects. So let's take this character I animated in Pixverse and get it to say something. I will upload the video, click lip sync. Hi, my name's Kevin. Welcome to Futurepedia. Now I'll select a voice and click generate voice. Hi, my name's Kevin. Welcome to Futurepedia. I don't like that option so I'll click generate again. Hi, my name's Kevin. Welcome to Futurepedia. And I can do that until I have one I like. Then attach and continue, then generate. And here's what we got. Hi, my name's Kevin. Welcome to Futurepedia. Here's one of the ones from the character example I did. Hi, my name's Kevin. Welcome to Futurepedia. Now let's generate some sound effects. I'll upload this explosion video. Click sound effects. I'll just write explosion. It generates three options to choose from. I like this one, so continue. Then attach to video. Easy as that. Under the free plan, you get 250 credits when you sign up, then 30 free credits per day. Each generation costs 10 credits, so once you run out of those first ones, it's only three generations per day. But the lip sync feature only uses two credits each, so you can use that quite a bit for free. Then generating sound effects doesn't use any credits at all. Now I'm going to put all of these tools together for the finale. I took a character I animated in Vigl, put it over the animations I got from Pixverse, Added some sound design and lip syncing and pica. I created the song for it in Udio, which again is also free. Here's what I got. It's a perfect day. Oh, such a perfect day. So poetic. Now, honestly, I had a whole storyline I was working on with that. but it just was not coming together and it felt super cheesy so I tossed most of it and added the dancers to salvage it into a joke. Alright, make sure to check out futurepedia.io to stay informed about everything in AI. You can search tools, save favorites to your profile, get recommendations, and also there's a bunch of new features added to the site like a curated list of tutorials on specific use cases, this AI innovations tracker that updates daily with all the AI advancements from the top 100 tech companies. It's a really easy way to get the information in a concise manner instead of scrolling through clickbait headlines and Twitter posts. So yeah, there's a bunch of new stuff there to check out and stay informed. You can sign up for the newsletter while you're there to get tools, tips, and tutorials delivered straight to your inbox every week as well. Thank you so much for watching. I'll see you in the next one.

You can replace any character with a single input image, and it's super easy. There's endless memeable use cases, a lot have gone viral already, but there's more practical ways to use it too. So let's jump in.

I'm in Discord right now, you go to any of these animate channels, and there's two main commands to use this, either with their preset videos or inputting your own. I'll start with the presets, for that we use forward slash animate. Then I'll upload a picture of myself.

Then if you use this link right here it will take you to their site with all sorts of preset videos to choose from. Tons of dance moves. You got the Joker.

Here's the Lil Yachty clip. I'll try this little dance out. Copy that, paste it in here, then select white or green if you want to remove the background or template to keep the one that's already there and use fine tune on, then submit. Depending on the time of day these may come back in 30 seconds or a few minutes. Then it will look something like Really solid and it's so easy to do.

But I want to use my own video instead of a preset. For this, we'll use forward slash mix. Upload that same image. Now I can upload my own video. So this is a short clip from Dance of the Sugar Plum Fairy.

I want to use that because, you know, I never took the time to learn ballet. I'll do green so the background is easy to remove and fine tune on. This is what came back which is amazing considering how complex some of the movement is. The model is able to guess all the angles that aren't shown in the input image.

Same with if you just use an image of your upper body, it will guess on the legs and feet too. It's not perfect yet, but pretty incredible. Now I need to bring this into a video editor. I'm using Premiere, this could all be done in a free editor like DaVinci Resolve, or even a simpler one like CapCut.

I've got the original footage on the timeline. Now I'll drag the green screen footage from Vigl over the top. Then go to Effects and use Ultra Key. Drag that onto the clip to apply it.

Then up to the Effect Control Panel, click on the eyedropper next to key color, then click green anywhere on the clip and it's gone. But you can still see the original dancer underneath. To fix that we'll go to runway ml and use the inpainting tool. There's a ton of other tools in here but this is all we need for this.

Upload the video and I'll drag that on. Now I'll paint over the character I want removed. Now it thinks for a second then fills in the background. And just from that one keyframe, it will guess what it's supposed to do across the whole video.

This shot isn't too complex, so it does it really well. And I can export this. Under the free plan, you can only export up to 720p. That's what this clip was already. But also we're going to upscale this later anyways.

Now I can bring that into Premiere and replace the original clip. And it looks great. And a quick tip in Premiere, since the edges are a little hard and don't blend well, Back on the effect control panel, you can click matte cleanup and add a little bit of softness to smooth it out.

That helps sell it a little more. So I'll export that. Then I want to upscale to increase the resolution. The best and easiest free upscaler I've found is right in CapCut. It's under magic tools, then video upscaler.

Just upload. Upscale 2x takes just a little bit and now we have our high definition clip all finished. Let's check it out. Alright that's amazing. I think I'd make a beautiful ballet dancer.

And I know everyone out there is probably asking, how would that look if you were wearing a leotard? Well, here you go. Just beautiful.

That's probably too much of a tangent to show that full process here, but for anyone wondering, I uploaded myself as a character reference in Midjourney, added a style reference I found, then face swapped that image using Insight Face Swap. Some of the outtakes trying to get that image were Pretty horrifying. Not sure if I should even show some of them on YouTube.

Alright, Vigl is a ton of fun, but I'll show some other limitations to save you some time. Adding a tutu looked funny but has no real like bouncing physics in the fabric. Another area that's difficult is fast moving scenes like this one.

It's really difficult for Runway to inpaint it with a constantly changing background like that. There's other solutions but not easy or free. Also if there's something that covers up part of your body in the video, like right here where the chain covers the arm and the wrecking ball covers part of the leg. That would involve some kind of complex masking to fix. Plus this had the same issues with the moving background.

There's endless other ways you could use Vigl. It works with all sorts of images. They could be cartoons, Pixar style, or here's an example of one with an illustration style generated with Firefly.

I could go on and on messing around with Vigl. It's a ton of fun, but we've got to move on. The fun doesn't stop there. If you want to learn more about taking advantage of AI for your YouTube channel or business, I highly suggest you check out this ebook, Using Generative AI to Scale Your Content Operations.

I've put a link in the description for you to download it for free. This is a super comprehensive guide that covers different types of AI tools, how to write prompts for them, and many ways to use AI like to simplify your research, and Breakthrough Writer's Block, which I use it for a lot. I really like that there are specific prompts written out, and I found it particularly helpful that they added some things to watch out for.

like bias, risks of plagiarism, and they clarify how Google handles AI content. It's always good to be aware of the limitations and pitfalls as well. This ebook was made by HubSpot, which is today's video sponsor.

Big shout out to them for this free resource. Next I want to show another really new free tool called Noisy. You can input the URL of a song, then add a text prompt, and optionally add an image reference, then it will create an entire animated music video.

It's still in Discord, but they're launching the web version any day now. So go check to see if it's available there first, depending on when you're watching this video. I've got a song I generated in Udio, which is another free tool. My last video was all about it. This is a metal song about gargoyles.

Let's go in for a Dio vibe. I'm going to use this as a reference image. I'll need a URL for it. This process will be much easier when the site launches, but this is the process right now. So I go down to this channel and send it here, then copy the link and save that in a notepad.

Now here's my song in Udio. I copy that URL. This also works with Suno, Soundcloud, YouTube, or you could send any MP3 in that same channel and use that URL. Now I'll use forward slash imagine.

I'll paste the Udio URL. Now I'll copy that image URL I saved and paste it here. Then add my prompt. Gargoyles atop cathedrals flying through the night.

Hit send and wait a couple minutes for it to generate. Once it's done you can click the edit button and it will take you to the site where you can regenerate any clip. I usually just like to generate a couple full versions of the song, then mix and match the clips. Let's see what we got.

That is awesome. It's pretty crazy how easy it is to get that whole video. I made a bunch of other styles with Noisy in my last video, which was that deep dive into Udio. Here's a couple clips from that to give a feel of the different styles it can do. I'll just play little bits of each though.

There is a house way down in New Orleans I'll call you sweetheart again, even if you drove your car right off that overpass. Now I'll jump into Pixverse. It does text-to-video, image-to-video, and also can do consistent characters.

They do have a web app for a better UI, but they have limited generations per day there. Right now you can still generate as many as you want for free in Discord. At least, I haven't hit a limit.

So I'll start with text-to-video, which is forward slash create. I'll do... A waterfall flowing into a pool surrounded by trees.

Then there's an option to select style. I'll use realistic. Then the aspect ratio 16 9. I'll actually copy the prompt so I can run it in each style for a comparison.

And submit. It looks like anime was done first. It comes back with four options. I'll click through each. Those all turned out really well.

Then I can either upscale my favorite or run a variation of it to get more similar options to choose from. Let's go check out Realistic. These are good...

Not great. Now 3D animation. Those all look great too.

The only style it really struggled with was realistic. Different styles work better depending on the prompt. Let's move on to Image to Video. Forward slash, Animate. I've got this nuclear explosion I generated in Midjourney, and I'll drop it in here.

Then you can select how much motion you want. It defaults to 5. I'll just leave it there. You can add a prompt. Explosion. Then I'll switch to HD quality, which makes it take a little longer.

So while that's working, I'll get another one started. The first one is done. And that looks awesome.

So I can click right here to upscale it. Let's check out the other one. That looks great too.

Alright, now I'll jump onto the site to show the character feature. You could do this in Discord too. If you click create, you can do all the stuff I just showed.

There's even some extra controls for camera movement. But we'll jump over to the character tab. So click create character.

Then I'll upload an image of myself as a Pixar character and name it... Kevin. Now when I click create, then click character, I have the option to choose from any character I've uploaded, which is only this one right now.

Then I just type a prompt and submit. I'll try a few different ones. So here's all of them next to each other.

You can see it's a very consistent character across all the shots. There's another free text-to-video and image-to-video tool called Hyper. I want to show that really quick.

It's in Discord and uses the same commands as Pixverse. The downside is the generations are only two seconds. So I'll jump in here and use forward slash create, then run the same prompts. Nuclear explosion, a waterfall flowing into a pool surrounded by trees. Then I'll use forward slash animate and bring in this same image.

Let's check these out. So Hyper also does a really good job, but the two seconds is a pretty big limitation. Although if you get creative you can still come up with some cool stuff. I saw this one on Twitter today.

Next up I want to show how you can add lip syncing to any of your videos. and also generate sound effects. So let's take this character I animated in Pixverse and get it to say something.

I will upload the video, click lip sync. Hi, my name's Kevin. Welcome to Futurepedia.

Now I'll select a voice and click generate voice. Hi, my name's Kevin. Welcome to Futurepedia. I don't like that option so I'll click generate again. Hi, my name's Kevin.

Welcome to Futurepedia. And I can do that until I have one I like. Then attach and continue, then generate. And here's what we got.

Hi, my name's Kevin. Welcome to Futurepedia. Here's one of the ones from the character example I did. Hi, my name's Kevin.

Welcome to Futurepedia. Now let's generate some sound effects. I'll upload this explosion video.

Click sound effects. I'll just write explosion. It generates three options to choose from.

I like this one, so continue. Then attach to video. Easy as that.

Under the free plan, you get 250 credits when you sign up, then 30 free credits per day. Each generation costs 10 credits, so once you run out of those first ones, it's only three generations per day. But the lip sync feature only uses two credits each, so you can use that quite a bit for free.

Then generating sound effects doesn't use any credits at all. Now I'm going to put all of these tools together for the finale. I took a character I animated in Vigl, put it over the animations I got from Pixverse, Added some sound design and lip syncing and pica. I created the song for it in Udio, which again is also free. Here's what I got.

It's a perfect day. Oh, such a perfect day. So poetic. Now, honestly, I had a whole storyline I was working on with that.

but it just was not coming together and it felt super cheesy so I tossed most of it and added the dancers to salvage it into a joke. Alright, make sure to check out futurepedia.io to stay informed about everything in AI. You can search tools, save favorites to your profile, get recommendations, and also there's a bunch of new features added to the site like a curated list of tutorials on specific use cases, this AI innovations tracker that updates daily with all the AI advancements from the top 100 tech companies. It's a really easy way to get the information in a concise manner instead of scrolling through clickbait headlines and Twitter posts.

So yeah, there's a bunch of new stuff there to check out and stay informed. You can sign up for the newsletter while you're there to get tools, tips, and tutorials delivered straight to your inbox every week as well. Thank you so much for watching.

I'll see you in the next one.

Transcript for:Overview of Free AI Video Tools

Transcript for:
Overview of Free AI Video Tools