Loading summary
Transcript1 lines
- [00:00]
A
11 Labs has just released an API for their music generator. And this is something absolutely no one is talking about and I am super excited. So today on the podcast I'm going to tell you why and also exactly where I think we're going to start seeing this implemented into tons of software that we use today. Overall, I'm excited. I'm a music enthusiast, so as you know, this is, this is a good one for me. Before we get into that, I wanted to mention if you want to try all of the models that we talk about on the podcast, I love for you to check out AI Box AI. This is, this is my own AI platform that's currently in beta. We give you access to the top 40 different AI models for 20 bucks a month. Claude, OpenAI, Google and dozens more 11 labs for audio generation. A bunch of cool image generation models. 20 bucks a month and you get access to all of it. So go check it out. AI Box AI. There is a link in the description. Let's get into what eleven Labs has announced today. So a bunch of this is coming from their Twitter. They recently had like an X thread and they said, introducing the 11 Labs Music API. This is the first music API for developers trained on licensed data and cleared for broad commercial use. Okay, this is why I'm excited about it. Personally, I'm actually someone that has used a ton of different image generate or audio music generation models. The number one, that model that I typically use is Suno AI. It's specifically for music. There's a bunch of cool stuff and I have used this. I've released full albums on Spotify. They get thousands of daily streams with AI music. I know, I don't know, it's like controversial. Some music artists will come at me. I put my own special touch. There's all sorts of cool things you can do with suno. For example, if you think, think of a song, you can hum a tune into it and then you can say basically the style. You want this to be a folk song and it will generate it. You can add lyrics or you can do instrumentals, which is what I typically do a lot. I'm not an amazing vocalist. And so yeah, you can do all sorts of cool things and make them really your own. So anyways, um, that's what I've typically used. Now the problem is it is just a consumer tool. There is no API. You can't tie this into a platform or a software or tool. And so really it kind of limits the reach unless you're signing up on their actual website. You're not going to be able to basically get the capabilities everywhere else. That has all sorts of interesting use cases which I'll talk about. So 11 Labs has changed that. One of the number one audio platforms of all time is 11 labs for creating vocal or for, you know, voices and reading and just basically audio. And now they've added music. Just recently they added their music and so they've added now the API which is really cool. I was actually excited when I saw their music thing was the first time I was like, oh, I should probably look at something other than Suno because the quality on it looked really good. And then of course now it has an API. It like, honestly, it's possible they jumpstart suno. It's kind of crazy to think about. Mid Journey was like the number one image generator, image generator for like a long time. And I almost never hear anyone talk about it anymore. The quality might be good but like the use is going to be super small because they basically lock themselves inside of Discord for a long time. And the number one thing that I think is a problem is that they've never allowed an API. So something like me on, on AI box, AI, my own platform, like I have OpenAI's image newest image generation model because they release an API and we get thousands of images generated from people that want to try that out and use it there. And there's no way for me to add Discord without doing like a shady third party API which could get broken. And just basically no software companies are going to touch it without an API and most of them can't. So anyways, this is really exciting. They said you can now integrate the highest quality AI music into your products and workflows. Since launch, creators have generated over 750,000 songs with 11 Labs music. This just came out like less than a month ago, I want to say. Um, and yeah, this is impressive. Almost a million songs have been generated since that happened. I am not surprised. 11 Labs has a huge user base and now that they basically have this API, I expect to see that number explode. So, honestly, something I'm really excited about. Um, this is what they specifically said. They said the 11 Labs Music API allows you to generate high quality tracks from a text prompt. And the reason I'm bringing this up is because like when an API comes out, it kind of makes sense for something like ChatGPT. ChatGPT API. You ask it questions, it responds. But when you do like music or video or even images, like there's a whole bunch of capabilities that sometimes the, the foundational, like the company that created it have, but they don't always move over to the API. So they outline basically what you're able to do with the API. Suno for example, lets you hum a tune and it will change it into a song. But if they had an API, I would assume like a feature like that wouldn't get ported over. Not necessarily, you know, for sure, but things like that don't typically, or maybe they come later in any case. So this is kind of what eleven Labs is outlining. So they basically let you put in a prompt, generates a tract. You can create vocals or instrumental versions of any genre or in any genre, which is really cool. You can customize the length, structure and the language. So they do have a whole bunch of really interesting settings there. And basically all these settings are what you can expect in a whole bunch of different software coming soon. And they've announced specifically specific partnerships they have already made, basically who they use to test and create this API, which I thought was cool. So they said this is created in collaboration with labels, publishers and artists. Songs were created with 11 labs music API are available for broad commercial use. Again, amazing because of all of the lawsuits with Suno and Yuio and basically sooner in Udio, like if you have their paid plan, they say you can use it for commercial use. They have like lawsuits based off of their training data with I think like different record labels and stuff. But I think that's kind of going to fall on them. They say it's good for commercial use. So I think basically you can use it. But 11 Labs is actually going quite a step further. When they did their training data I noticed one of my distributors that I use is called Lander and inside of Lander there was this pop up that came, I want to say six months ago, maybe even a year ago, which basically started saying like, hey, would you like to opt in to have your music included into AI training? We will compensate you for it basically. And, and some artists I said yes, some artists I said no on this. And it's interesting because basically when I saw this I was like okay, cool. I thought maybe it was Suno or Udio they were trying to like a lot of these AI companies are like train on data. They weren't allowed to. And then they go back and try to like get licensing for data and do a retrain and then when the lawsuits are all happening they're like oh look, whatever, we just, we retrained and now we're, we got the good data anyway. So this is kind of isn't important. Let's move past it. I feel like that's kind of like OpenAI and the new York Times, like that's kind of like a strategy. So I thought that was, I was assuming that was the what was happening. But actually beyond Lander just building another revenue stream for themselves. Eleven Labs, I'm assuming, probably even contacted Lander and a lot of the distributors to try to work out these deals, make that pop up, show up. Because Lander's data they did confirm was embedded in this model and all those people are opting in. So it's like, you know, for anyone that's like, oh, you know, these AI models are training off of people's music and then they're, you know, replacing artists and stuff. Well, if you're using something from 11 labs, all of these artists have opted in and they're getting compensated. So my mind, like this is zero percent controversial and just a new revenue stream for artists. Okay. They said it's designed for building apps across media and entertainment, whether you're delivering personalized meditations, producing music for video games, or creating AI generated ads. Over the past weeks, we worked closely with partners like VEED amp and Hedra Labs to build the 11 Music API. Okay, I want to talk about the partners, but first I got to talk about a very funny element and that is the fact that they said based off of software. One of the, one of like the startup use cases they mentioned here was personalized meditations, which is interesting because a lot of distributors, lender in particular, who has this partnership giving them data, actually don't allow you to release like meditation music onto Spotify. Some people do, some people don't. Lander specifically doesn't, which is interesting because basically Lander's data was used and then the use case that they're recommending you can actually publish onto Spotify. But I guess if you're just doing software like on inside of an application, then you can do it. So anyways, kind of an interesting thing. So they made a bunch of partners. The one that I'm the most excited about, they have amp, they have Hedger Labs. Hedger basically does videos, which is really, really, really cool. They have a, a demo where basically it's like a woman singing an 80s song. And so Hedger makes the video of the woman singing, but now they actually have the song to go with it and it's lip syncing up with the video. That is very, very cool for me, in my opinion, moving beyond just making videos, but now you're actually making a music video, which is crazy. Another one of their partnerships, which I think is pretty big is with VEED. VEED is, you know, this huge video generation tool. It, it sponsored like a lot of podcasts a couple of years ago. I think it sponsored the this Week in Startups podcast. And so like, I just know VEED has been shilled to me like over 100 episodes that I've listened to. But it's actually a great video generation or video editing platform. I think it's basically a competitor, Capcut, kind of like Canva for video. I think in any case it has a ton of users and so one of the big use cases that they're saying people are going to be using it for is, is AI background music. Right? So if you're generating, if you're making a video, which is something that I personally have, you know, experienced a lot where I'm generating like some sort of video or editing a video for some like a YouTube project. This is, you know, I used to have a YouTube channel where I'd put out weekly YouTube content before I had an editor and you know, people that helped with the production on my podcast. And yeah, I remember this was a huge thing that would take me so much time. Basically whenever I wanted like some B roll and wanted to put music in the background, I go to like YouTube has this like free music library place and like you could like search through and, and like pick different songs. And it would take like, I'm ashamed to say I would sometimes spend an hour finding like the right music for my video because I'd just be listening to it and like, you know, if you had a good song, you just kind of like use it over and over again. You'll notice this, like Mr. Beast. Mr. Beast or Mark Robert, like Mark Rober has the same song and all of his B roll and he probably got that custom made so that he doesn't have to pay licensing for it. So like that's probably one element of it. But also then he doesn't have to sit there and think about what, what the newest B roll song is going to be. So anyways, I think this basically solves that problem where you're generating, you're creating and editing a video and now you can generate music in the background. You just describe the style that you want, you can do it. And basically they say you can create a custom track. The AI analyzes your videos transcripts to generate a track that perfectly matches the subject and tone of your content. To me, that is Very cool. You don't have to explain what you want. You just say, hey, generate like music here. It's going to look at the transcripts of your whole video and it's like, oh yeah, basically let's make a song like X, Y, Z. Now I wonder if I'm hoping you can switch. Actually they did say in their API you have the ability to switch between making it an instrumental or with vocals, but either way, very, very cool option, they said. To showcase the power of 11 music in our API, our team also built Jingle Maker, a tool to create songs based on your X profile or website in just two clicks. Uh, this is actually really funny. They got the, the website jinglemaker AI and I think that more companies should 100% do this. And basically if you just, if you go to Jinglemaker AI, which I did and, and tested it out because I thought this was hilarious, you could put a website or you could put your, your profile in. Now the one thing I'll say is there, it's sort of like misleading on how it happened or the using your own profile. But basically the way they explain it is like you, you upload the link to your, your Twitter account, your X profile and it's gonna make like a song. Now my assumption was it's gonna be like a song about me based off my content. I did that. But it just made a song about like X in general. So I don't think it like actually looked at the full URL. Anyways, whatever, I'll forgive them for it. It was pretty funny. It was a good song. You can make a whole bunch of different styles of music and it's a very simple website. You basically can make a radio ad, Midwest farmer, a rap drama trailer, US politician style song, whatever. So like there's all these like funny things that you can basically make. Now I think more companies should do this because basically they're displaying like they're showing, look, here's a new product and we've embedded the API in here and here's how it works and they make this kind of viral loop where if it actually, okay, it would be more viral if you could actually make a song about your personal profile and then you probably think it's really funny and post it. Unfortunately it doesn't look like that actually works. So I imagine the viral, the virality of that is not going to hit the way they hope unless they fix that. I don't know. But in any case, I think this is genius. Make a viral loop with a product that's super funny. And it basically shows off what your product is capable of. I think more companies should do this. So I was excited. They said their API is available at 50% off this month so if you want to start try using it now. It's going to be 50% off. And yeah, I think this is amazing. Really, really cool tool. Eleven Labs is, you know, the leader in voice, so it's cool to see them getting in more with actually making music, which is a growing industry. Anyways, thank you so much for tuning in to the AI Chat podcast. If you enjoyed the episode today, make sure to leave a rating and review wherever you get your episodes and also make sure to go check out AI Box AI if you want to try all of the AI models I talk about on the show in one place for one subscription price. You don't have to have subscriptions to 10 different platforms, you get them all on the AI box website. Thanks so much for tuning in and I will catch you in the next episode.