Kits.AI - 100 Years of Audio

It hasn't been 100 years since we launched Kits.AI, but it has been one year—and as of today, we have processed over 100 years of user audio on the platform. We owe this achievement to the 5 million creators and 50,000 Kits customers who helped us get here. With this milestone, we’re taking a moment to reflect on our journey in the wild world of AI music.

Cutting Through the AI Hype

In the midst of viral ‘AI cover’ headlines last summer, we set forth with a belief that AI music tech will be an unlock for serious musicians with real problems. As one of the first AI singing products to market we had an influx of users with wildly different use cases and intent, which made it difficult to know exactly who we were building for. After some time (and many user conversations), we found that the ones consistently picking up the phone to tell us how Kits saved them time and money were music producers and vocalists.

Producers using Kits voice clones for virtual sessions with artists touring halfway across the world, artists who lost the ability to sing restoring the voice of their youth on Kits and countless other stories inspire us today. We’re only in the first inning of building AI tools for producers, and we can’t wait to hear stories from our creators in the next year of Kits.

Research at Kits AI

Even in our earliest days, it was clear that audio quality of outputs on Kits is critical to your experience. This is why AI Audio research has been a top priority for us this past year, driving significant improvements in audio quality and realism when compared to other solutions. We’ve made big strides from where we started, and we’ll be sharing more detailed insights into our advancements in AI singing soon.

Research not only drives quality, but unlocks the music workflows of tomorrow. Creatives are increasingly recognizing the paradigm shift generative AI tools represent for music and are adopting them more and more every day. We’re excited to push the boundaries of what’s possible in AI music and we'll be sharing more about our plans for the next generation of tools in the coming months.

How We Build

A lot has changed for us and the AI music space in the past year, but one thing remains constant: building AI music technology that works for everyone—artists, producers, and right holders alike—is a huge challenge. Last year’s headlines of ‘Spotify Removes AI Drake Songs’ have turned to ‘Labels Sue AI Music Companies’ and tensions between building great products and respecting the work of artists are at an all time high.

Our mission to help musicians has always gone deeper than what we build, and very much into how we build. We’ve put massive amounts of time and energy into sourcing and licensing every second of Kits training data from real musicians with real compensation. As we train the AI audio models of the future, we will continue to build them with all parties in mind. For more information on Kits data practices, see our ethics page.

It’s been an incredible first year helping musicians create better music faster, and we’d like to thank every producer, vocalist, and creator who has used Kits.

Here’s to the next 100 years of audio,

Evan