Yes, it’s possible but not perfect. You’d need software that uses AI or spectral editing to separate stems (like guitar, vocals, drums, etc.). Examples include iZotope RX, Moises.ai, or Lalal.ai. The challenge is streaming sources—usually you need to download the audio file first, since most tools require a local file to process.