Back to home

About YuE2 & yue2music.com

Independent cloud studio powered by YuE2 open weights. Learn about symbolic planning architecture and our mission.

Last updated: 2026-09-19

About YuE2 & yue2music.com

Independent Studio Notice: yue2music.com is an independent third-party web creation studio and cloud runtime powered by the open-source YuE2 foundation model. We are not affiliated with, endorsed by, or representing Multimodal Art Projection (M·A·P), HKUST, Stanford, or any university institutions. All trademarks, model names, and paper copyrights belong to their respective academic authors.

What is YuE2?

YuE2 (Music with a Plan) is an open-weights foundation model for full-song music generation developed and published by Multimodal Art Projection (M·A·P) and academic collaborators (including HKUST, Tokenwave.AI, NYU, Stanford, and MBZUAI).

Unlike traditional black-box generators, YuE2 introduces Symbolic Planning: it drafts an editable musical score (ABC format) before rendering audio, giving musicians transparent control over harmony, melody, and structure.

What is yue2music.com?

Running YuE2 locally requires dedicated high-VRAM GPUs (e.g. NVIDIA A100/H100/RTX 4090), Linux environments, and specialized Python dependencies.

yue2music.com bridges this gap by providing an accessible, browser-based cloud studio:

  • Zero Local Setup: Run YuE2 inference pipelines on managed GPU clusters with standard web browsers.
  • Interactive Score Editor: Inspect ABC notation, audition chords with Web Audio synths, and customize lyrics before initiating neural audio generation.
  • 48kHz Master Audio: Deliver high-fidelity stereo outputs rendered through flow matching and YuE2-VAE.

Foundation Model Architecture

YuE2's official architecture consists of:

  • YuE2-3B: The primary full-song generation model unifying symbolic planning with flow matching diffusion.
  • YuE2-VAE: A high-fidelity 48kHz stereo neural audio autoencoder with low reconstruction artifacting.
  • SheetSage2: An all-in-one audio-to-score transcription model capable of transcribing audio into editable ABC notation.
  • MERT2 & MERT2-FS: Self-supervised music representation models providing acoustic embeddings.
  • WildSongBench: An open evaluation benchmark of 192 diverse prompts for fair comparisons.

Academic Research & Foundation Model Credits

We gratefully acknowledge the original authors and institutions behind the open-weights research:

  • Multimodal Art Projection (M·A·P)
  • The Hong Kong University of Science and Technology (HKUST)
  • Tokenwave.AI
  • New York University (NYU)
  • Stanford University
  • Mohamed bin Zayed University of Artificial Intelligence (MBZUAI)

Our Mission at yue2music.com

Our goal is to make music creation transparent, controllable, and democratized. We believe artificial intelligence should empower human creativity rather than replace it. By combining the precision of musical theory with the expressive power of neural audio diffusion, we provide artists with an intelligent co-creator that speaks the language of music.

Whether you are writing a jazz ballad, scoring an indie game, experimenting with microtonal scales, or laying down cyber-rock guitar riffs, YuE2 gives you the steering wheel.

Contact & Community