# LeVo: High-Quality Song Generation with Multi-Preference Alignment

arXiv: https://arxiv.org/abs/2506.07520 · HTML: https://mlsanity.com/paper/2506.07520/

## Key facts

- Cited by 25 papers in the map (data through 14 Oct 2026).
- Placed in Preference alignment for recommenders and RAG, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 55 of its 65 references that are in the map.
- First posted on arXiv 9 Jun 2025.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Preference alignment for recommenders and RAG](https://mlsanity.com/alignment/preference-alignment-for-recommenders-and-rag/)

## Builds on (most cited first, 30 of 55)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [High-Resolution Image Synthesis with Latent Diffusion Models](https://mlsanity.com/paper/2112.10752/) (2021)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Scalable Diffusion Models with Transformers](https://mlsanity.com/paper/2212.09748/) (2022)
- [Qwen2.5 Technical Report](https://mlsanity.com/paper/2412.15115/) (2024)
- [Neural Discrete Representation Learning](https://mlsanity.com/paper/1711.00937/) (2017)
- [Robust Speech Recognition via Large-Scale Weak Supervision](https://mlsanity.com/paper/2212.04356/) (2022)
- [wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations](https://mlsanity.com/paper/2006.11477/) (2020)
- [Taming Transformers for High-Resolution Image Synthesis](https://mlsanity.com/paper/2012.09841/) (2020)
- [Scaling Autoregressive Models for Content-Rich Text-to-Image Generation](https://mlsanity.com/paper/2206.10789/) (2022)
- [LaMDA: Language Models for Dialog Applications](https://mlsanity.com/paper/2201.08239/) (2022)
- [SoundStream: An End-to-End Neural Audio Codec](https://mlsanity.com/paper/2107.03312/) (2021)
- [Neural Codec Language Models are Zero-Shot Text to Speech Synthesizers](https://mlsanity.com/paper/2301.02111/) (2023)
- [FastSpeech 2: Fast and High-Quality End-to-End Text to Speech](https://mlsanity.com/paper/2006.04558/) (2020)
- [Vector-quantized Image Modeling with Improved VQGAN](https://mlsanity.com/paper/2110.04627/) (2021)
- [AudioLM: a Language Modeling Approach to Audio Generation](https://mlsanity.com/paper/2209.03143/) (2022)
- [Qwen2-Audio Technical Report](https://mlsanity.com/paper/2407.10759/) (2024)
- [Jukebox: A Generative Model for Music](https://mlsanity.com/paper/2005.00341/) (2020)
- [Simple and Controllable Music Generation](https://mlsanity.com/paper/2306.05284/) (2023)
- [MusicLM: Generating Music From Text](https://mlsanity.com/paper/2301.11325/) (2023)
- [High-Fidelity Audio Compression with Improved RVQGAN](https://mlsanity.com/paper/2306.06546/) (2023)
- [AudioLDM 2: Learning Holistic Audio Generation with Self-supervised Pretraining](https://mlsanity.com/paper/2308.05734/) (2023)
- [WhisperX: Time-Accurate Speech Transcription of Long-Form Audio](https://mlsanity.com/paper/2303.00747/) (2023)
- [NaturalSpeech 2: Latent Diffusion Models are Natural and Zero-Shot Speech and Singing Synthesizers](https://mlsanity.com/paper/2304.09116/) (2023)
- [Fast Timing-Conditioned Latent Audio Diffusion](https://mlsanity.com/paper/2402.04825/) (2024)
- [MuseGAN: Multi-track Sequential Generative Adversarial Networks for Symbolic Music Generation and Accompaniment](https://mlsanity.com/paper/1709.06298/) (2017)
- [Meta Audiobox Aesthetics: Unified Automatic Quality Assessment for Speech, Music, and Sound](https://mlsanity.com/paper/2502.05139/) (2025)
- [Hybrid Transformers for Music Source Separation](https://mlsanity.com/paper/2211.08553/) (2022)

## Built on by (most cited first, 25 of 25)

- [DiffRhythm 2: Efficient and High Fidelity Song Generation via Block Flow Matching](https://mlsanity.com/paper/2510.22950/) (2025)
- [JAM: A Tiny Flow-based Song Generator with Fine-grained Controllability and Aesthetic Alignment](https://mlsanity.com/paper/2507.20880/) (2025)
- [Muse: Towards Reproducible Long-Form Song Generation with Fine-Grained Style Control](https://mlsanity.com/paper/2601.03973/) (2026)
- [Multi-Metric Preference Alignment for Generative Speech Restoration](https://mlsanity.com/paper/2508.17229/) (2025)
- [SongBench: A Fine-Grained Multi-Aspect Benchmark for Song Quality Assessment](https://mlsanity.com/paper/2604.25937/) (2026)
- [LeVo 2: Stable and Melodious Song Generation via Hierarchical Representation Modeling and Progressive Post-Training](https://mlsanity.com/paper/2606.30642/) (2026)
- [SongPrep: A Preprocessing Framework and End-to-end Model for Full-song Structure Parsing and Lyrics Transcription](https://mlsanity.com/paper/2509.17404/) (2025)
- [CMI-RewardBench: Evaluating Music Reward Models with Compositional Multimodal Instruction](https://mlsanity.com/paper/2603.00610/) (2026)
- [Pushing the Frontier of Full-Song Generation: Hierarchical Autoregressive Planning Meets Flow-Matching Rendering](https://mlsanity.com/paper/2607.20253/) (2026)
- [EvoAudio: Recursive Self-Improvement for Audio Understanding](https://mlsanity.com/paper/2609.27389/) (2026)
- [Towards Hallucination-Free Music: A Reinforcement Learning Preference Optimization Framework for Reliable Song Generation](https://mlsanity.com/paper/2508.05011/) (2025)
- [ImmersiveFlow: Stereo-to-7.1.4 spatial audio generation with flow matching](https://mlsanity.com/paper/2601.12950/) (2026)
- [Shao: Scaling Acoustic Token Language Models Toward High-Fidelity Music Generation](https://mlsanity.com/paper/2605.01790/) (2026)
- [Anysynth:Zero-Shot Instrument Cloning via In-Context Learning and Asymmetric Hierarchical Guidance](https://mlsanity.com/paper/2607.11143/) (2026)
- [Tutti: Expressive Multi-Singer Synthesis via Structure-Level Timbre Control and Vocal Texture Modeling](https://mlsanity.com/paper/2602.08233/) (2026)
- [LoSATok: Low-Dimensional Semantic-Acoustic Tokenizer for Cross-Domain Audio Understanding and Generation](https://mlsanity.com/paper/2605.27840/) (2026)
- [SegTune: Structured and Fine-Grained Control for Song Generation](https://mlsanity.com/paper/2606.02638/) (2026)
- [Phun-Bench: Evaluating LLMs on Phonological Understanding in Chinese](https://mlsanity.com/paper/2606.07300/) (2026)
- [MADB: A Large-Scale Music Aesthetics Dataset with Professional and Multi-Dimensional Annotations](https://mlsanity.com/paper/2607.06929/) (2026)
- [MusicMark: A Robust Generative Watermarking Framework for Music Generation](https://mlsanity.com/paper/2607.11117/) (2026)
- [WanSong v1.0 Technical Report](https://mlsanity.com/paper/2607.14749/) (2026)
- [MuseCritic: Learning Multi-Aspect Song Rewards through Natural-Language Aesthetic Critiques](https://mlsanity.com/paper/2608.11755/) (2026)
- [AURA: Unified Multimodal Framework for Conversational Music Editing](https://mlsanity.com/paper/2609.14344/) (2026)
- [YuE2: Unifying Symbolic and Audio Music Generation at Frontier Quality](https://mlsanity.com/paper/2609.33757/) (2026)
- [AdaLoop: Adaptive-Depth Latent Reasoning for Audio Language Models](https://mlsanity.com/paper/2610.06949/) (2026)

## Abstract (arXiv)

Recent advances in large language models (LLMs) and audio language models have significantly improved music generation, particularly in lyrics-to-song generation. However, existing approaches still struggle with the complex composition of songs and the scarcity of high-quality data, leading to limitations in audio quality, musicality, instruction following, and vocal-instrument harmony. To address these challenges, we introduce LeVo, a language model based framework consisting of LeLM and Music Codec. LeLM is capable of parallel modeling of two types of tokens: mixed tokens, which represent the combined audio of vocals and accompaniment to achieve better vocal-instrument harmony, and dual-track tokens, which separately encode vocals and accompaniment for high-quality song generation. It employs two decoder-only transformers and a modular extension training strategy to prevent interference between different token types. To further enhance musicality and instruction following ability, we introduce a multi-preference alignment method based on Direct Preference Optimization (DPO). This method handles diverse human preferences through a semi-automatic data construction process and post-training. Experimental results demonstrate that LeVo significantly outperforms existing open-source methods in both objective and subjective metrics, while performing competitively with industry systems. Ablation studies further justify the effectiveness of our designs. Audio examples and source code are available at https://levo-demo.github.io and https://github.com/tencent-ailab/songgeneration.