Hugging Face trending modelspublished Sep 9, 18:30 UTC (6d ago)ingested Sep 11, 04:16 UTCm-a-p

Part of a story covered by 3 sources: “YuE2 open music model tops Suno v5 on SongBench; DeepSeek ships 552B DeepSeek-V4.1-Flash with 1M-token context” — merged summary and timeline →

m-a-p/YuE2-3B — new model trending #30 on Hugging Face

infoModel releaseimportance 28

AI summary · glm-5.3-flash

M-A-P released YuE2-3B, an open music generation model that outperforms Suno v5 on WildSongBench and runs locally on a 24GB GPU.

The M-A-P (multimodal-art-projection) team released YuE2-3B, an open-weights music generation model that turns lyrics and a style prompt into full songs with vocals and accompaniment. It uses an AR-NAR Mixture-of-Transformers backbone with symbolic planning and flow matching through a VAE, and supports editable scores (melody and chords, including ABC notation) plus agentic editing workflows. On 192 WildSongBench prompts it reports a SongBench average of 6.9632 (best-of-8) versus 6.8721 for Suno v5, claimed as state of the art among evaluated open and proprietary models. It runs 48 kHz stereo inference locally on a single 24GB NVIDIA GPU without quantization, with companion releases including YuE2-Vae, MERT-v2 encoders, the WildSongBench dataset, and SheetSage2.

Open-weights model generates complete songs with vocals from lyrics and style prompts.
Editable score workflow supports melody/chord planning in ABC notation and agentic revisions.
Reports SongBench average 6.9632 on WildSongBench, ahead of Suno v5's 6.8721.
Runs 48 kHz stereo inference locally on a single 24GB GPU without quantization.
Companion releases include YuE2-Vae, MERT-v2 encoders, WildSongBench, and SheetSage2.

Read the original

VendorsHugging Face M-A-P

ProductsYuE2-3B YuE2-Vae MERT-v2 SheetSage2 Suno v5

OrganizationsMultimodal Art Projection (M-A-P)

AI modelsYuE2-3B

Tagsmusic-generation open-weights text-to-audio m-a-p yue2 suno wildsongbench mixture-of-transformers

Full article2,363 words · extracted from huggingface.co · click to collapse

Frontier music generation with editable scores

<a href="#quick-start">🚀 Quick start</a>

<a href="#cover-an-existing-song">🎙️ Cover</a>

<a href="#speed-and-resources" title="Speed and resources">⚡ Speed</a>

<a href="#benchmarks">📊 Benchmarks</a>

<a href="#citation">📚 Citation</a>

**YuE2 is an open music generation model that rivals Suno v5.** Turn lyrics and a style prompt into a complete song with vocals and accompaniment, then shape its melody and chords through an editable score.

**State-of-the-art results on WildSongBench.** YuE2 (best-of-8) achieves the highest SongBench average among all evaluated open and proprietary models: **6.9632**, compared with **6.8721** for Suno v5.

![YuE2 song quality and text alignment on WildSongBench](assets/figure1.png)

*Frontier song quality and text alignment on 192 WildSongBench prompts. YuE2 uses symbolic planning; Bo8 means best-of-8.*

- **Compose and edit:** melody + chords, melody-only, or direct generation; bring your own ABC score.

- **Edit with an agent:** turn musical feedback into score, style and lyric revisions, then let YuE2 render the next version. [Hear the editing process](https://map-yue2.github.io/#agentic-music-editing).

- **Run locally:** 48 kHz stereo songs on a 24GB GPU, without quantization.

- **Build on it:** Hugging Face loading, text guidance (CFG), and separate planning and synthesis APIs.

![YuE2 architecture](assets/architecture.png)

*One AR–NAR Mixture-of-Transformers backbone writes the score and semantic tokens, then generates acoustic latents through flow matching. The VAE turns them into stereo audio.*

## 🎧 Listen to YuE2

### 🎶 Text-to-music

Original songs generated from lyrics and a style prompt.

**Cyber Metal · English · 5:00**

**今晚不眠 · Mandarin funk / nu-disco · 3:24**

**Passion · English rock · 3:55**

*All three songs use [🤗 YuE2-Vae](https://huggingface.co/m-a-p/YuE2-Vae).*

### 🎙️ Cover songs

Existing songs reimagined in a new style.

[Make your own cover →](#cover-an-existing-song)

**Auld Lang Syne · Jazz-funk cover · 3:10**

**最炫民族风 · Ballad cover · 4:45**

**Jingle Bells · Heavy metal cover · 1:09**

### 🤖 Agentic editing

**[Explore the agentic editing demo →](https://map-yue2.github.io/#agentic-music-editing)**

Follow **The Last Train** through **9 steps and 14 versions**, from Mandarin pop to English jazz with modern harmony and a saxophone solo built around two complete statements of “Twinkle, Twinkle, Little Star.” Hear the full songs and inspect the conversation, scores, prompts and lyrics at each step.

[Try editing with an agent →](#export-a-plan-edit-it-and-generate)

## 🚀 Quick start

Linux · Python 3.10+ · 24GB NVIDIA GPU with BF16 support. Install the inference package:

```bash

python -m pip install huggingface-hub==0.36.2

hf download m-a-p/YuE2-3B yue2_infer-0.1.5-py3-none-any.whl --local-dir .

python -m pip install ./yue2_infer-0.1.5-py3-none-any.whl

```

[Create](#generate-a-song) · [Cover](#cover-an-existing-song) · [Edit & agentic edit](#export-a-plan-edit-it-and-generate)

Load the pipeline once for the examples below:

```python

from pathlib import Path

from yue2 import YuE2Pipeline

pipe = YuE2Pipeline.from_pretrained("m-a-p/YuE2-3B", device="cuda")

```

### 🎶 Create

Turn a style prompt and lyrics into a complete song with vocals and accompaniment.

Use the [style and full lyrics from 今晚不眠](examples/tonight-awake.json), the funk / nu-disco demo above.

```python

import json

from huggingface_hub import hf_hub_download

repo = "m-a-p/YuE2-3B"

prompt_path = hf_hub_download(repo, "examples/tonight-awake.json")

demo = json.loads(Path(prompt_path).read_text(encoding="utf-8"))

style, lyrics = demo["style"], demo["lyrics"]

song = pipe(style=style, lyrics=lyrics, cot="full", seed=demo["seed"])

song.save("song.flac")

song.save_artifacts("outputs/song") # ABC, tokens, latents, audio and settings

```

Defaults are ready to use: `cot="full"` and [🤗 YuE2-Vae](https://huggingface.co/m-a-p/YuE2-Vae).

| Option | What it does |

|---|---|

| `cot="full"` | Melody + chord planning (default) |

| `cot="melody"` | Melody-only planning; recommended for covers |

| `cot="off"` | Generate without a symbolic plan |

| `cfg_scale=1.2` | Experiment with stronger text guidance |

### 🎙️ Cover

Start from an existing recording and give it a new arrangement.

**For covers, we recommend melody-only mode (`cot="melody"`).**

1. **Get the score:** transcribe the existing song with [🤗 SheetSage2](https://huggingface.co/m-a-p/SheetSage2) and save its melody ABC **without chord symbols** as `melody.abc`.

2. **Get the lyrics:** ask an agent to find them online, or transcribe the singing with [Qwen3-ASR](https://huggingface.co/Qwen/Qwen3-ASR-1.7B) or the [Gemini API](https://ai.google.dev/gemini-api/docs/audio). Check the words, organize them into sections matching the recording, and save them as `cover_lyrics.txt`.

Text extracted automatically; images, tables and formatting may be missing. Original: https://huggingface.co/m-a-p/YuE2-3B