MusicLang” is the contraction of “Music” & “language”: we aim to bring advanced controllability features and high-quality music generation by manipulating symbolic music.

Our vision:

Music, as a universal language and a common cultural good, inherently belongs to everyone. Adopting an open approach to music generation aligns with this ethos, fostering inclusivity and collaboration within the creative community. Transparent access to code and algorithms empowers collaboration among musicians, researchers, and developers, advancing both the technical capabilities of music AI and fostering diverse artistic expression.

Our product:

MusicLang foundation model: our foundation model for creating and generating original midi soundtracks;
MusicLang predict: our AI prediction api of the MusicLang package;
MusicLang Language: a new language for tonal music. This language allows composers to load, write, transform and predict symbolic music in a simple, condensed and high level manner;
MusicLang Demo Space: all the above in action, to offer advanced controllability features and high-quality music generation by manipulating symbolic music. Also available in our Colab.

Technical details:

If you want to learn more about how we are moving toward symbolic music generation, go to our technical blog. The tokenization, the model are described in great details:

What for?

🎶 You can, for example, enforce a chord progression on the generated music;
🎹 Generate a music from scratch that you can export to your favorite DAW in MIDI format;
🕺 Use your own music as a template to create new composition to extend an existing track;
🪩 Fast & reliable: We target fast inference on CPU aiming to distribute it in DAWs & plugins;
And more.

For whom?

MusicLang is well suited for:

R&D engineers working on music generation use cases: we have come up with a SOTA model to bring advanced controllability features over music generation;
For technical creators: with a bit of technical knowledge, you can add a brick to your composition workflow. MusicLang is well suited if you want to generate music that you can export to your favorite DAW in MIDI, generate a chord progression with a full control on the desired chord progression, time signature & more;
For musicians: adjust your own creations endlessly until you discover the perfect variation for your project.

We are looking for contributors to help us improve the model, the tokenization, the performances and the documentation. Whether you're contributing code or just saying hello, we'd love to hear about the work you are creating with MusicLang. Here's how you can reach out to us:

Join our Discord to ask your questions and get support
Follow us on Linkedin
Add your star on GitHub or HuggingFace