Music creation for the next generation: Arpeggi Studio & Kits
Music creation for the next generation: Arpeggi Studio & Kits

Subscribe to Arpeggi Labs

Subscribe to Arpeggi Labs

Arpeggi Labs raises $5.1m led by a16z crypto to build open-source music creation
As the world becomes increasingly digital, our ability to collaborate and build on the ideas of others has never been easier. Technology has led to a creative explosion for many artistic mediums, but today’s outdated music infrastructure hinders artists from the natural creative process of sharing music for reuse and iteration. Tracking down rights holders, papering legal agreements, exploiting licensing fees… We started Arpeggi Labs to create a better way to make music. Arpeggi is building w...

A Year in Review
A year ago today, we set out to revolutionize the way music is created, transformed, and shared online. Music-making has always been collaborative, but recent advancements in cryptographic technology have opened up new possibilities for decentralized, permissionless collaboration over the internet, and that’s when Arpeggi comes in. We created a brand new DAW, built natively on-chain, to pull from and contribute to a freely-licensed sound library where attribution is automatically tracked and ...

Kits AI: Your Voice, Your Terms
The State of AI Voice ModelsOver the last few months, AI voice conversion technology has led to an explosion of music made in the likeness of some of the world’s most popular artists, from Drake and The Weeknd to Grimes. This new technology has the potential to be one of the most revolutionary tools in music history, but it presents major challenges for artists.Artist control: Unofficial, unlicensed AI voice models let anyone say anything in an artist’s voice, with no way for artists to regul...

Arpeggi Labs raises $5.1m led by a16z crypto to build open-source music creation
As the world becomes increasingly digital, our ability to collaborate and build on the ideas of others has never been easier. Technology has led to a creative explosion for many artistic mediums, but today’s outdated music infrastructure hinders artists from the natural creative process of sharing music for reuse and iteration. Tracking down rights holders, papering legal agreements, exploiting licensing fees… We started Arpeggi Labs to create a better way to make music. Arpeggi is building w...

A Year in Review
A year ago today, we set out to revolutionize the way music is created, transformed, and shared online. Music-making has always been collaborative, but recent advancements in cryptographic technology have opened up new possibilities for decentralized, permissionless collaboration over the internet, and that’s when Arpeggi comes in. We created a brand new DAW, built natively on-chain, to pull from and contribute to a freely-licensed sound library where attribution is automatically tracked and ...

Kits AI: Your Voice, Your Terms
The State of AI Voice ModelsOver the last few months, AI voice conversion technology has led to an explosion of music made in the likeness of some of the world’s most popular artists, from Drake and The Weeknd to Grimes. This new technology has the potential to be one of the most revolutionary tools in music history, but it presents major challenges for artists.Artist control: Unofficial, unlicensed AI voice models let anyone say anything in an artist’s voice, with no way for artists to regul...
Share Dialog
Share Dialog
>300 subscribers
>300 subscribers


Last week, we announced that Kits was developing a solution for artists to safely license their voice. Tomorrow, we’re releasing a first look into our solution with artist TK, live on Kits.io.
We’ve worked with artists to solve licensing protection, control over derivatives, and upfront monetization for AI Voice Models
TK will be releasing the first official Kits voice model on June 1st at 10am PST
We will be releasing more models on the kits.ai platform over the coming weeks
One of the most exciting things about AI Voice Models is the ability to crowdsource creative content with an artist's voice and likeness. With this opportunity for collaborative upside also comes unique challenges around control and misuse of artist’s voices. Through our conversation with artists, we’ve identified three key problem areas in AI Voice Cloning:
Artist licensing: Streaming services have been forced to take down an influx of unlicensed AI music, begging the question: how can AI voice conversion technology be licensed to publish compliant releases?
Artist control: Unofficial, unlicensed AI voice models let anyone say anything in an artist’s voice, with no way for artists to regulate content and ensure their likeness is used how they want it to be used.
Artist monetization: Artists have no way to monetize the usage of their AI voice models or capture royalties from works created with their voice and likeness.
Today’s release with TK tackles these challenges head on. Here’s a look into how our solution puts artists in the driver’s seat.

Licensing provides an important framework for protecting artists and ensuring they capture value when their songs are used in movies or remixed by other artists. After extensive experience working with artist advisors and legal counsel to adapt the Can’t Be Evil Licenses for Kits Sample Packs, we’ve developed a brand new license for AI Voice Models that puts the artist first.
With our AI Voice Model license, users who transform their voice receive a personal-use license but must receive approval from the artist for any commercial release of their derivative song. This system allows for artist voice models to be accessible to users while ensuring artists retain control over their commercial likeness. Artists set their own royalty splits for derivative works.
Beyond requiring approval for any commercial releases, our AI Voice Model License comes with a hate-speech termination clause, allowing the artist to terminate the license of any derivatives made in their voice that they disagree with.
AI Voice Models have the opportunity to be a massive new income stream for artists, but today’s solutions completely cut artists out. With our solution, artists can monetize upfront by charging for access to their model and capture backend royalty splits on approved derivative works using their voice.
TK’s model is a flat fee for 30 days of access, but in the coming weeks, we’ll transition our platform to pay artists as their models are used by implementing a platform credit system. With kits.ai we’re building faster than ever to ensure that the control of this powerful new technology is put in the hands of artists.
If you’re an artist interested in submitting vocals for a model or simply learning more, fill out this form. To stay up to date, subscribe here.
Last week, we announced that Kits was developing a solution for artists to safely license their voice. Tomorrow, we’re releasing a first look into our solution with artist TK, live on Kits.io.
We’ve worked with artists to solve licensing protection, control over derivatives, and upfront monetization for AI Voice Models
TK will be releasing the first official Kits voice model on June 1st at 10am PST
We will be releasing more models on the kits.ai platform over the coming weeks
One of the most exciting things about AI Voice Models is the ability to crowdsource creative content with an artist's voice and likeness. With this opportunity for collaborative upside also comes unique challenges around control and misuse of artist’s voices. Through our conversation with artists, we’ve identified three key problem areas in AI Voice Cloning:
Artist licensing: Streaming services have been forced to take down an influx of unlicensed AI music, begging the question: how can AI voice conversion technology be licensed to publish compliant releases?
Artist control: Unofficial, unlicensed AI voice models let anyone say anything in an artist’s voice, with no way for artists to regulate content and ensure their likeness is used how they want it to be used.
Artist monetization: Artists have no way to monetize the usage of their AI voice models or capture royalties from works created with their voice and likeness.
Today’s release with TK tackles these challenges head on. Here’s a look into how our solution puts artists in the driver’s seat.

Licensing provides an important framework for protecting artists and ensuring they capture value when their songs are used in movies or remixed by other artists. After extensive experience working with artist advisors and legal counsel to adapt the Can’t Be Evil Licenses for Kits Sample Packs, we’ve developed a brand new license for AI Voice Models that puts the artist first.
With our AI Voice Model license, users who transform their voice receive a personal-use license but must receive approval from the artist for any commercial release of their derivative song. This system allows for artist voice models to be accessible to users while ensuring artists retain control over their commercial likeness. Artists set their own royalty splits for derivative works.
Beyond requiring approval for any commercial releases, our AI Voice Model License comes with a hate-speech termination clause, allowing the artist to terminate the license of any derivatives made in their voice that they disagree with.
AI Voice Models have the opportunity to be a massive new income stream for artists, but today’s solutions completely cut artists out. With our solution, artists can monetize upfront by charging for access to their model and capture backend royalty splits on approved derivative works using their voice.
TK’s model is a flat fee for 30 days of access, but in the coming weeks, we’ll transition our platform to pay artists as their models are used by implementing a platform credit system. With kits.ai we’re building faster than ever to ensure that the control of this powerful new technology is put in the hands of artists.
If you’re an artist interested in submitting vocals for a model or simply learning more, fill out this form. To stay up to date, subscribe here.
No activity yet