YuE2: Open-Source Music Model From Launch to Free Commercial Use

Episode 1 · Open-source music model YuE2-3B launches with editable scores, claims Suno v5-level quality (2026-09-10, 12 posts)

M-A-P released YuE2-3B, an open-weight music generation model now trending on Hugging Face, with a live demo at map-yue2.github.io. Given lyrics and a style prompt, it generates complete songs up to 5 minutes with vocals and accompaniment at 48kHz stereo, runs locally on a 24GB GPU, and supports Chinese and English for creation, cover, and editing workflows. Early testers report impressive sound, though the non-commercial license limits commercial use.

Confirmed

• The model is roughly 3B parameters (some sources cite 3.59B, 28 layers, AR/NAR expert architecture), with an accompanying arXiv paper and tags including music-generation and agentic-editing
• Its "white-box" pipeline first produces an editable symbolic plan of melody and chords in ABC notation that users can inspect, play, and edit before rendering the final song
• The team claims YuE2 (best-of-8) beats Suno v5 on WildSongBench; other accounts place it between Suno 4.5 and 5.5, surpassing Mureka v9 and MiniMax 3 on several benchmarks
• Community reactions: @realmrfakename found the output excellent and comparable to Suno v5/v5.5 but was disappointed by the non-commercial license; @wzwowzw0002 asked about ComfyUI integration; @bdsqlsz questioned whether it truly beats Suno V5, awaiting community testing; @Acceptable-Cycle4645 called the demo quite good

Unconfirmed

• Claims of matching or beating Suno v5/v5.5, Mureka v9, and MiniMax 3 come mainly from the developers and individual listeners; broader independent evaluation is still lacking

Why it matters

• Editable scores make generation transparent and controllable, offering musicians a "plan first, render later" workflow unlike most end-to-end black-box models
• Local 48kHz stereo full-song generation on a 24GB consumer GPU significantly narrows the gap between open-source and closed commercial systems like Suno
• The non-commercial license remains its biggest limitation, leaving practical commercial prospects unclear

• Open-source YuE2 music model generates editable symbolic scores before rendering full songs — GreyScope · 2026-09-10
• Open-Weight Music Model YuE2-3B Released, Claimed to Rival Suno v5.5 — realmrfakename · 2026-09-10
• YuE2-3B: Open Music Model With Editable Scores Beats Suno v5 on WildSongBench — realmrfakename · 2026-09-10
• Open-weight music model YuE2-3B debuts: up to 5 minutes, claims parity with Suno v5.5 — realmrfakename · 2026-09-10
• Open music model YuE2-3B beats Suno v5 on SongBench, runs locally on 24GB GPU — realmrfakename · 2026-09-10
• Open-source music generator YuE2 tops SongBench, beating Suno v5 with symbolic planning — solyarisoftware · 2026-09-10
• YuE2 combines symbolic planning and audio generation, claims Suno v5-level music quality — wzwowzw0002 · 2026-09-10
• Open-source YuE2 generates full songs via symbolic planning, taking on Suno V5 — bdsqlsz · 2026-09-10
• Open-source 3B music model Yue2-3B released, author asks if it beats Suno V5 — bdsqlsz · 2026-09-10
• m-a-p releases YuE2-3B, an open music generation model trending on HF — m-a-p · 2026-09-10
• YuE2-3B open-source music generation model released with demos — Acceptable-Cycle4645 · 2026-09-11
• Open-source music model YuE2 claims to beat Mureka v9 and MiniMax 3, near Suno 4.5-5.5 — _AustinCalvert_ · 2026-09-11

Episode 2 · Open-Source Music Model YuE2 Beats Suno v5/v6 on WildSongBench (2026-09-14, 4 posts)

M-A-P released YuE2, a 3B open-source music generation model that runs locally on a 24GB GPU and tops WildSongBench, beating closed models like Suno v5/v6 and marking a new milestone for open music AI.

• Open music model YuE2 (3B) beats Suno v5/v6 on WildSongBench, runs on a 24GB GPU — AdinaYakup · 2026-09-14
• YuE2-3B open music model lands on Hugging Face, tops WildSongBench over Suno v5/v6 — AdinaYakup · 2026-09-14
• Music model YuE2 trends on Hugging Face, finetuned from SheetSage2 — Comfy-Org · 2026-09-15
• Open-source YuE2 beats Suno v5/v6 on WildSongBench with symbolic score planning — solyarisoftware · 2026-09-15

Episode 3 · YuE2 Music Model Gets ComfyUI Nodes and LoRA Trainer (2026-09-15, 2 posts)

Community developers released ComfyUI-FL-YuE2, bringing the open-source YuE2 music generation model into ComfyUI with lyrics-to-song support and in-workflow LoRA training based on an RVQ Adapter.

• ComfyUI LoRA Trainer Released for YuE2 Music Model — thatisnotmychapstick · 2026-09-15
• YuE2 music generation lands in ComfyUI with editable piano roll and AR LoRA training — PurzBeats · 2026-09-15

Episode 4 · Open-Source Music Model YuE2 Runs Suno-Quality Songs Locally in Just 4.5GB VRAM (2026-09-15, 7 posts)

YuE2, an open-source music generation model with only 3B parameters, has been tested by multiple developers and is being called possibly the strongest open-source option today: high quality, fast, with extremely low VRAM requirements — it runs locally on consumer GPUs and even some laptops, with cover quality approaching closed-source services like Suno. Pinokio author cocktailpeanut has integrated it into WanGP, dramatically lowering the deployment barrier.

Confirmed

• cocktailpeanut integrated YuE2 into WanGP and monitored VRAM throughout: generation never exceeded 4.5GB VRAM, with output at 48kHz stereo, enabling "SUNO-grade" songs on consumer hardware including some laptops.
• YuE2 is available on Wan2gP with one-click install via Pinokio.
• A French developer tested the cover feature: feeding YuE2 a 30-year-old personal composition plus lyrics produced a dozen high-quality covers in varied styles, quickly.
• @lazyspock installed the not-yet-stable ComfyUI support and built a workflow: the INT8 CONROT quantized version needs about 8GB VRAM (runs on an RTX 4070 12GB), generating a four-minute song in 120–150 seconds, with covers close to Suno.
• One tester called it "one of the craziest things I've seen," marveling at its low-VRAM performance; cocktailpeanut noted YuE-2 has just 3B parameters.

Unconfirmed

• One repost about "VRAM never exceeding 4.5GB" lacked full context identifying the tested model, though the figure matches cocktailpeanut's own YuE2 testing.

Why it matters

• A 4.5GB VRAM floor means ordinary users can generate music locally without high-end GPUs, and with Pinokio one-click install the deployment cost is near zero.
• For the first time, an open-source model poses real competition to closed services like Suno in both quality and ease of use, while local operation avoids copyright and paywall constraints.

• YuE2 tested: the first real local Suno rival, near-Suno covers on 8GB VRAM — lazyspock · 2026-09-15
• Dev measures under 4.5GB VRAM peak throughout full local generation — cocktailpeanut · 2026-09-16
• Open-source YuE2 generates SUNO-quality songs with just 4.5GB VRAM — cocktailpeanut · 2026-09-16
• Open-source music model YuE2 runs locally on just 4.5GB VRAM, one-click via Pinokio — cocktailpeanut · 2026-09-16
• YuE2 music model runs locally on 4.5GB VRAM: one-click install, batch style covers from a single track — cocktailpeanut · 2026-09-16
• Open-source music model YuE2 runs locally on 4.5GB VRAM, outputs 48kHz full songs — cocktailpeanut · 2026-09-17
• YuE-2 music generation model stuns users with quality at just 3B parameters — cocktailpeanut · 2026-09-17

Episode 5 · YuE2 allows individuals to commercialize generated music for free (2026-09-16, 2 posts)

The YuE2 team confirmed that individuals may use the open-source music model's generated content commercially for free, while commercial licensing applies only to enterprises, lowering the barrier for independent creators.

• YuE2 music model: individual creators can use generations commercially, license only applies to companies — cocktailpeanut · 2026-09-16
• YuE2 team confirms individuals can use generations commercially, license only for companies — wzwowzw0002 · 2026-09-16