YuE2: the 3B open-source music model punching above its weight
renderartist · reddit · 2026-09-17
Reddit user renderartist shared a hands-on test of the open-source music model YuE2, finding that this tiny 3B-parameter model punches far above its weight, with mostly clean vocals — a rare free local alternative for music generation.
What the author did:
- Trained a small LoRA on a handful of niche 80s tracks using ai-toolkit, with impressive results
- Used an LLM to quickly build a threejs visualizer and render the generated music into a video
- Captioning still has kinks, and a refinement method is needed
Having been disappointed by ACE Step, the author was skeptical at first but now expects YuE2 to pick up steam in the coming weeks.
More from Multimodal
- A 25-year design veteran asks: who still mixes pro software with AI tools? — Opening-Knee-5913 · 2026-11-02
- Google Pay pilots AI fashion Discover tab in India with virtual try-on and photo search — heyshrutimishra · 2026-11-02
- Reddit user vents on MiniMax: plasticky AI look, 20GB+ models eating SSD space — DelboyTrigger · 2026-11-02
- Paper folk-story prompt template produces charming cut-out collage images — azed_ai · 2026-11-02
- SeedVR2 vs MiniMax Open-H3 vs FlashVSR+ Turbo: real 4K upscale timings on RTX 5090 — CeFurkan · 2026-11-02
- Could one person with Kling and CapCut tell the same story? — tetsuoai · 2026-11-02