Back to directory

YuE2

An open-source music generation model that plans a song's melody and chords symbolically before rendering a complete track with vocals and accompaniment from lyrics and a style prompt.

YuE2

YuE2 is a music generation model that unifies symbolic and audio music generation. Given lyrics and a style prompt, it first writes a melody-and-chord plan, then renders that plan into a complete song with vocals and accompaniment.

Because the composition is planned symbolically, a person or an agent can read, play and edit the melody and chords before rendering, and the same generation checkpoint supports zero-shot covers that reimagine a transcribed song in a new style, along with agentic editing that revises a song's score, arrangement or lyrics through conversation. On the WildSongBench benchmark, its best-of-8 setting reports the highest mean score among evaluated systems, including proprietary music generation systems.

An earlier version, YuE, remains available on a separate branch with its own code, documentation and license. YuE2 is written in Python and released under the Apache-2.0 license.

More in Generative media

See all 146
  • Blender Copilot

    github.com

    A Copilot-style AI chat panel inside Blender that runs its agent loop in Blender's own Python process, executes the model's generated code against the live scene, and wraps every conversational turn in a single undo step.

  • A ComfyUI add-on that patches ComfyUI-GGUF so Qwen-Image-2.1 models can run fully as GGUF files, fixing a text-encoder mismatch and missing architecture detection until upstream support arrives.

  • A ComfyUI node that expands a short prompt into a full styled prompt, negative prompt and aspect ratio for Qwen-Image-2.1, using one of 17 photographic styles.