YuE2 · Frontier Music with Symbolic Planning
The article details advancements in AI-driven music generation and analysis, specifically highlighting the performance of YuE2 and MERT2 models on industry benchmarks. These tools demonstrate significant progress in symbolic planning, melody transcription, and audio quality.
Why it matters
Showcases the rapid evolution of generative AI in creative fields and the development of standardized benchmarks for evaluating synthetic audio.
Listen to a song, then explore the melody, rhythm, and chords in its symbolic plan.
A familiar song can take a different shape. Listen to changes in melody, lyrics, tempo, and arrangement.
A song takes shape through a conversation. Loading the editing story…
The listening selection, gathered across genres and languages.
YuE2 (best-of-8) reaches 6.9632 on SongBench, the highest observed mean among 15 evaluated settings on WildSongBench (192 prompts). Suno v5 scores 6.8721 in the same comparison.
September 5, 2026 evaluation · rankings vary by metric.
WildSongBench. 192 prompts and 15 system settings. The table reports automatic evaluation scores. Best-of-8 selects one of eight generations by musicality, prompt control, and lyric accuracy.
Figure 1. Song quality combines SongBench and SongEval; text alignment combines MuLan, AllMusicCaps, and prompt control. Both axes show normalized comparison indices. Bubble area represents AudioBox production quality.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in