Awesome Music Generation Papers 🎵

Curated list of groundbreaking music generation research.

Text-to-Music Generation 📄➡🎸

• MusicGen Simple and Controllable Music Generation [2024] [Meta] [Paper] [Code]

• MusicLM: Generating Music From Text [2023] [Google] [Paper]

• InspireMusic: Integrating Super Resolution and Large Language Model for High-Fidelity Long-Form Music Generation [2025] [Alibaba] [Paper] [Code] [Demo]

• FluxMusic FLUX that Plays Music [2024] [Skywork] [Paper] [Code]

• Magenta RT Live Music Models [2025 [Google] [Paper]

Lyrics-to-Song Generation 📜➡🎶

• ACE-Step 1.5: Pushing the Boundaries of Open-Source Music Generation [2026] [Paper] [GitHub]

• HeartMuLa: A Family of Open Sourced Music Foundation Models [2026] [PKU] [Paper] [GitHub]

• Muse: Towards Reproducible Long-Form Song Generation with Fine-Grained Style Control [2026] [FudanNLP] [Paper] [GitHub]

• LeVo: High-Quality Song Generation with Multi-Preference Alignment [2025] [Tencent] [Paper] [GitHub] [Demo]

• SongBloom: Coherent Song Generation via Interleaved Autoregressive Sketching and Diffusion Refinement [2025] [CUHK-Shenzhen] [Paper] [Demo] [GitHub]

• ACE-Step: A Step Towards Music Generation Foundation Model [2025] [GitHub]

• YuE: Scaling Open Foundation Models for Long-Form Music Generation [2025] [m-a-p] [Paper] [Code] [Demo ]

• DiffRhythm: Blazingly Fast and Embarrassingly Simple End-to-End Full-Length Song Generation with Latent Diffusion [2025] [ASLP-lab] [Paper] [Code] [Demo] [HuggingFace]

• SongCreator: Lyrics-based Universal Song Generation [Paper] [Demo]

Controllable Music Generation & Editing ⚙️

• MuseControlLite: Multifunctional Music Generation with Lightweight Conditioners [2025] [Paper]

• Editing Music with Melody and Text: Using ControlNet for Diffusion Transformer [2024] [ICASSP] [Paper]

• Music ControlNet: Multiple Time-Varying Controls for Music Generation [2024] [Demo]

• Instruct-MusicGen: Unlocking Text-to-Music Editing for Music Language Models via Instruction Tuning [2024] [Paper]

• MusiConGen: Rhythm and Chord Control for Transformer-Based Text-to-Music Generation [2024] [Demo]

• MusicMagus: Zero-Shot Text-to-Music Editing via Diffusion Models [2024] [Paper]

• MelodyEdit: Zero-shot Music Editing with Disentangled Inversion Control [2025] [Paper]

• SongEditor: Adapting Zero-Shot Song Generation Language Model as a Multi-Task Editor [Paper] [Demo]

Singing Voice Synthesis (SVS) 🎼➡🎤

• DiffSinger: Singing Voice Synthesis via Shallow Diffusion Mechanism [2021] [Paper] [Code]

• VISinger 2: High-Fidelity End-to-End Singing Voice Synthesis Enhanced by Digital Signal Processing Synthesizer [2022] [Paper] [Code]

• TCSinger: Zero-Shot Singing Voice Synthesis with Style Transfer and Multi-Level Style Control [2024] [Paper] [Code]

• TCSinger 2: Customizable Multilingual Zero-shot Singing Voice Synthesis [2025] [Paper]

• SoulX-Singer: Towards High-Quality Zero-Shot Singing Voice Synthesis [2026] [Paper] [Code]

Singing Accompaniment Generation (SAG) 🎤➡🎸

• FastSAG: towards fast non-autoregressive singing accompaniment generation [2024] [Paper] [Code]

• SingSong: Generating musical accompaniments from singing [2023] [Google] [Paper]

• Streaming Generation for Music Accompaniment [2025] [MIT] [Paper]

• AnyAccomp: Generalizable Accompaniment Generation via Quantized Melodic Bottleneck [2026] [Paper]

Lyrics-to-Melody Generation (L2M) 📜➡🎼

• CSL-L2M: Controllable Song-Level Lyric-to-Melody Generation Based on Conditional Transformer with Fine-Grained Lyric and Musical Controls [2025] [AAAI] [Paper] [Code] [Demo]

• SongComposer: A large language model for lyric and melody composition in song generation [2024] [Paper] [GitHub]

Text-to-Symbolic Music Generation 📄➡🎼

• NotaGen: Advancing Musicality in Symbolic Music Generation with Large Language Model Training Paradigms [2025] [Tsinghua] [Paper] [Code] [Demo ]

• Mupt: A generative symbolic music pretrained transformer [2024] [m-a-p] [Paper] [Demo]

Music SSL & Encoder & Codec 🎶➡🔢

• MuQ: Self-supervised music representation learning with mel residual vector quantization [2025] [Tencent] [Code]

• MusicFM A foundation model for music informatics [2024] [Paper] [Code]

• MERT: Acoustic music understanding model with large-scale self-supervised training [2023] [m-a-p] [Paper] [GitHub]

• Mucodec: Ultra low-bitrate music codec [2024] [Tencent] [Paper]

Music Transcription 🎶➡🎼

• YourMT3+: Multi-Instrument Music Transcription with Enhanced Transformer Architectures and Cross-Dataset STEM Augmentation [2024] [Paper] [GitHub]

• Perceiver TF Multitrack music transcription with a time-frequency perceiver [2023] [ByteDance] [ICASSP] [Paper]

• MT3: Multi-task multitrack music transcription [2021] [Paper]

Music Source Separation 🎶➡🎹🎸🥁

• SCNet: Sparse compression network for music source separation [2025] [ICASSP] [Paper]

• Music source separation with band-split rope transformer [2024] [ICASSP] [Paper]

• Music source separation with band-split RNN [2023] [TASLP] [Paper]

Music Evaluation Metrics 📏

• SongEval: A Benchmark Dataset for Song Aesthetics Evaluation [2025] [ASLP-lab] [Paper] [GitHub] [Dataset]

• MusicEval: A Generative Music Corpus with Expert Ratings for Automatic Text-to-Music Evaluation [2025] [AISHELL] [Paper] [Dataset]

• Frechet Music Distance: A Metric For Generative Symbolic Music Evaluation [2024] [Paper] [GitHub]

Dataset & Benchmark 📊

• Marble: Music audio representation benchmark for universal evaluation [2023] [m-a-p] [Paper] [GitHub]

• CHORDONOMICON: A Dataset of 666,000 Songs and their Chord Progressions [2024] [Paper]

• FMA: A Dataset For Music Analysis [GitHub] [100K tracks]

Music Agent 🤖

• ComposerX: Multi-Agent Symbolic Music Composition with LLMs [2024] [Paper]

• AutoMV: An Automatic Multi-Agent System for Music Video Generation [2025] [Paper] [GitHub]

• MuseAgent-1: Interactive Grounded Multimodal Understanding of Music Scores and Performance Audio [2026] [Paper]

Website & Products 🌐

• Lamucal - Real-time audio to chords, lyrics, beat, and melody with Hybrid-Net [Web] [GitHub]
• Dango.ai - Any instruments separation and vocals extraction [Web]

Miscellaneous 🎲

• consonance-ACE From Discord to Harmony: Decomposed Consonance-Based Training for Improved Audio Chord Estimation [Paper] [GitHub]