Freebeat

What if you could drop your music track into an app, enter a short prompt, and a few minutes later you get a full-on music video? The tool I'm going to show you today does exactly that. So, if you are a music creator, but you don't have the money to produce a music video or you are a content creator and you want some nice visuals for your video or you just want to have some fun, then check out this video. We will try this tool with different music styles, electro, classical music, music with and without lyrics, and we will see how well it handles these different genres and how good is the character consistency when we add people to our music video. We will also take a look at the other AI video editing tools this website offers, and they all use the latest AI models like Sora 2 and Veo 3. In fact, some of the B-roll in this intro you just saw and the background music was generated using this tool. My name is Florian Walt and this is the AI tool corner where I review the latest AI software to find out which ones can actually improve our lives and businesses.

Okay, the tool is called Freebeat. I will put the link into the video description below. You get a bunch of free credits when you sign up so you can try this out for free. And when you sign up over my link and decide to later buy a subscription, you get a 10% discount. Let's click on login to create our account. After logging in, we get redirected to this agent page, which is the flagship feature of Freebeat that we will take a look at first. When we click on explore now, here we can also see a bunch of other AI image and video editing tools that we will check out later. For now, we want to go back to the music video agent.

So this video agent takes a lot of work from you. It analyzes your music track and then it comes up with a video concept and it describes all the different scenes and then it goes ahead and generates video for all these scenes and stitches them together. Let's try this out together.

So the first track I want to use is this one here from the no copyright sounds YouTube channel. We can either upload a music file here or we can just insert the YouTube link directly. It loads the track and then we can select the segment that we want to use in our video. We can select the whole music video, but the longer the video is, the more credits it uses up. So, let's use a shorter clip. So, I want to use this section of the track. Yeah, you will hear the rest in a moment. Let's add a short prompt. Actually, let's try to keep it very simple for now and just enter make a video for this song. Couldn't be any simpler, right?

We can also put characters into our video. Up to two. We can either select one of these predefined ones like Pikachu and the Joker, but we can also upload an image and use, for example, ourselves. Let's try that. So, I uploaded an image of myself. And for the other character since it's a female voice, maybe let's pick a woman. Maybe this futuristic looking one here.

Then you select the aspect ratio. If you want to upload this to like TikTok, you can choose 9:16, but for YouTube, you want to pick 16:9. We can set the resolution to 1080p full HD. We can also add captions. Let's pick these orange ones here. Doesn't really matter. We select a style. I like to use photo realistic. We don't want to have a watermark, so we enable this. And we keep auto enabled because then the agent executes all the steps automatically and we barely have to do anything. Do we want to make this video public? Actually, I don't care. And then we click on create and the agent gets to work.

It analyzes our input like the music and the prompt. And then it does a bunch of steps. You don't even have to read this. It comes up with a summary with different scenes. Does a lot of reasoning as you can see, but we can just lean back and wait for this to finish. And with regular AI video generation models, you would have to do all these steps yourself. You would have to come up with the different scenes, describe them, generate elaborate prompts, make sure these prompts fit together, generate the single video clips, which for most models, I think, can only be up to like 12 seconds long or so, and then you have to stitch them together. But the agent does all of these steps for you automatically, and in the end, we get a full music video.

Now, one weird quirk of the tool is that sometimes it tells us that we have to refresh the page because the agent stops generating output here. But I also got an email that told me that the music video is ready. So, we basically just have to refresh this page with F5. And then there's our music video. And from what I see, this person actually looks like me. Definitely has the same head shape. Let's check this out. Stranded in the open, dried out tears of sorrow.

Well, this is me with long hair apparently. But the face does look like me, doesn't it? I would actually confuse this for my twin. And remember that we didn't insert a prompt. So, the AI came up with this video concept. We can fine-tune this. And we will try this with another music track in a minute.

There's the other character. Yeah, it's pretty short because I limit it to 33 seconds.

Let's try a longer video. So, I uploaded another copyright free track that I really like, Cradles. And this time, let's make it 90 seconds long. Again, I want to include characters since it's kind of a creepy song. Maybe let's pick the Joker and the Black Queen here. Again, 1080p 16:9. I didn't even pay attention to the captions in the last video. Did we have captions? Let's try this again. Style. Just for fun, let's pick something else this time. Lego blocks seems interesting. I'm curious how it handles the Joker as a Lego character. No watermark.

This time I add a prompt. The Joker and the Black Queen are dancing on a graveyard with skeletons around them. Let's send this. And this is what I was talking about. The AI agent generates all these different video clips which you would have to do manually with a regular AI video generation tool. That's what makes agents so powerful. They execute a lot of steps that you would have to do manually.

And our music video is ready. Let's check it out. World of make-believe. Kids screaming in the cradles. Profanities. I see dancing. Wow.

Yeah. In a photo realistic video, the lips actually sync with the text of the music, but since those are Lego figurines, we don't have any mouth movements. Ones who heard my cries and watched me weep. I love everything. Fire spreading all around my room. My world's so bright, it's hard to breathe. But that's all right. Take my eyes open to force reality.

No. Why can't you just let me eat my weight and glee? I live inside my own world of make-believe. Kids screaming in the cradles. Profanities. Some days I feel skinnier than all the other days.

Yeah, it's kind of cool, isn't it?

And since this was so much fun, let's try out one more. This time Tchaikovsky classic that I'm sure you have heard before. Let's try out this one. This time, let's try SpongeBob playing a piano. SpongeBob plays the piano while vibing to the music. Maybe let's try 9:16 ratio this time. So, we could post this on TikTok. For the style, I'm going to select Midnight Blue Zone. Not sure what that means, but let's try it out. I picked around 1 minute of the track this time. And let's send this off.

And our 9:16 video is ready. Let's check it out. Heat. Heat. This one is a bit laggier.

Okay. And there is a fourth one I want to show you that I already created earlier. This one here, which is another cool track with a break dancer. Let's check it out. Louder, stronger, better, louder. Stronger.

Yeah, if you're a music creator, I think this is a nice way to showcase your music on YouTube without having to pay tons for a music video.

All right, that's the music video agent. As I mentioned before, Freebeat also has a bunch of other AI video and image generation tools. Let's check out some of them.

So, for example, we have this AI video generator which just generates a single AI video clip rather than a full music video. And this uses the latest models right now. This is for example Sora 2, PixVerse, or Veo 3.1. And this is a nice way for you to use these different AI video models through one single app. One of the B-roll clips in the intro was actually generated using this tool. This one here. That's it. Nothing left.

And the prompt was something like guy sitting on a piano showing his empty wallet. You can also upload the starting image here. Then give it a prompt and modify it. For example, I took this slightly autistic looking image of mine and I wanted AI to turn this into a superhero that's going up in flames. And this is the result I got. Yeah. And I told it that I want to fly away at the end of the video. So, with a bit of creativity, you can generate a bunch of these different clips, stitch them together, and create a full little movie.

In this video, I wanted it to put Blastoise toys from Pokemon into a wrestling match. But these AI models don't allow you to use any trademarked characters. So instead, I wrote a giant blue turtle with cannons on its shell. And this is what I got.

That's a giant turtle in the ring. And it's blasting Tyler with those cannons. Look at that pressure. He's getting absolutely hosed down. He can't even stay on his feet.

Anyway, you can play around with this. You have a lot of inspiration here. If you like one of these clips, you can click on go create and it puts the prompt right here. Then select your model, the duration of the clip, and generate your video.

AI video effect is also cool because this has a bunch of predefined prompts and some of them have been viral on TikTok. For example, making two characters kiss or hug each other or this one here where a guy takes his shirt off and he suddenly looks like me. Again, you can upload a starting image.

For example, I tried this one here, which is a very old photo of my grandparents. This is my mother and this is my uncle. Here in the prompt, I told the AI to make this little girl, which is my mother, dance, and the parents hug each other. And this is the result I got. Looks super cool if you ask me. There are some weird elements in here. For example, I'm not sure what she's holding in her hands, but it's still very impressive, especially how the camera moves around.

Of course, these AI video models get better and better in the future, and Freebeat will get updates and use these latest models. So, if you watch this video 6 months into the future, it will probably not use Sora 2 and Veo 3 anymore because there are newer models out by then.

Here's another one where I'm hacking Jesus because I'm such a good boy. And this one here is a fail. So, I took this image with my mother I made in New York and I told the AI to make us hug to make me hug my mother. And this is what happened instead. There's some random dude coming up and hugging my mother. What the hell? Yeah, sometimes these AI models fail.

And you even have an AI music generator here. This uses models like Suno or one of these other ones. You can add your prompt here and generate a whole music track. And the music you heard in the intro was actually generated through this. And I only tried a single prompt and I already got a great result, which is this one here. Sounds very professional.

You can play around with them and check out these other tools here like character swap. We have a video upscaler. If you have a low-quality video and you want to make it full HD, you can generate ASMR videos and a lot of other things.

If you want to try out Freebeat, click the link in the video description and you get a 10% discount. And remember, you also get a bunch of free credits so you can try this out without paying anything. Then subscribe to the channel for more AI tool reviews in the future. And I hope I see you in the next video. Take care.