Rightsify's GCX Source Separation datasets provide high-quality, rights-cleared audio data for training and evaluating AI models focused on isolating individual instrumental and vocal tracks from mixed audio recordings. Our extensive library covers a wide range of genres, artists, and recording techniques, enabling developers and researchers to create innovative applications for music production, remixing, and audio manipulation