🌐 English | 简体中文 | 繁體中文 | Español | Français | 日本語

NeurIPS 2025

arXiv Online Demo Hugging Face ModelScope

If you find this project useful,
a star ⭐ on GitHub would be greatly appreciated!

ThinkSound is a unified Any2Audio generation framework with flow matching guided by Chain-of-Thought (CoT) reasoning.

PyTorch implementation for multimodal audio generation and editing: generate or edit audio from video, text, and audio, powered by step-by-step reasoning from Multimodal Large Language Models (MLLMs).

• 2025年09月19日 🎉 ThinkSound has been accepted to the NeurIPS 2025 Main Conference!
• 2025年09月01日 🔥 Our AudioCoT dataset is now open-sourced and available on Hugging Face!
• 2025年07月17日 🧠 Finetuning enabled: training and finetuning code is now publicly available, along with clear usage instructions to help you customize and extend ThinkSound with your own data.
• 2025年07月15日 📦 Simplified installation and usability: dependencies on PyPI for easy cross-platform setup; Windows .bat scripts automate environment creation and script running.
• 2025年07月08日 🔧 Major update: model lightweighted and optimized memory and GPU usage, now supports high-throughput audio generation at scale!
• 2025年07月01日 🔥Online demo on Hugging Face Spaces and ModelScope for interactive experience!
• 2025年07月01日 🔥Released inference scripts and web interface;
• 2025.06 🔥ThinkSound paper released on arXiv!
• 2025.06 🔥Online Demo is live - try it now!

• Any2Audio: Generate audio from arbitrary modalities — video, text, audio, or their combinations.
• Video-to-Audio SOTA: Achieves state-of-the-art results on multiple V2A benchmarks.
• CoT-Driven Reasoning: Chain-of-Thought reasoning for compositional and controllable audio generation via MLLMs.
• Interactive Object-centric Editing: Refine or edit specific sound events by clicking on visual objects or using text instructions.
• Unified Framework: One foundation model supports generation, editing, and interactive workflow.

ThinkSound decomposes audio generation and editing into three interactive stages, all guided by MLLM-based Chain-of-Thought (CoT) reasoning:

1. Foley Generation: Generate foundational, semantically and temporally aligned soundscapes from video.
2. Object-Centric Refinement: Refine or add sounds for user-specified objects via clicks or regions in the video.
3. Targeted Audio Editing: Modify generated audio using high-level natural language instructions.

ThinkSound Overview

Environment Preparation:

git clone https://github.com/liuhuadai/ThinkSound.git
cd ThinkSound
conda create -n thinksound python=3.10
conda activate thinksound
pip install thinksound
conda install -y -c conda-forge 'ffmpeg<7'
# Download pretrained weights https://huggingface.co/liuhuadai/ThinkSound to Directory ckpts/
# model weights can be also downloaded from https://www.modelscope.cn/models/iic/ThinkSound
git lfs install
git clone https://huggingface.co/liuhuadai/ThinkSound ckpts
# To improve inference and training speed, you may optionally install a FlashAttention backend compatible with your system and PyTorch version.

✅ Windows Tip:
Windows users can simply run setup_windows.bat (or double-click it) to automatically create the conda environment, install all dependencies (including FFmpeg), and download the pretrained model — no manual setup required.
Make sure conda and git are installed and available in your system PATH before running the script.

chmod +x scripts/demo.sh
./scripts/demo.sh <path-to-your-demo-video> <title> <CoT description> [use-half]

You can use the provided .bat script instead:

.\scripts\demo.bat <path-to-your-demo-video> <title> <CoT description> [use-half]

Note:

• <path-to-your-demo-video>: The path to a single video
• [use-half] (optional): Add use-half at the end to enable half precision feature extraction.

chmod +x scripts/eval_batch.sh
./scripts/eval_batch.sh <video_path> <csv_path> <save_path (optional)> [use-half]

Use the equivalent .bat script:

.\scripts\eval_batch.bat <video_path> <csv_path> <save_path (optional)> [use-half]

Note:

• <video_path>: Path to the root directory containing all .mp4 videos to be processed (all videos must be of equal duration).
• <csv_path>: A CSV file with text prompts for each video (see demo_test.csv for format).
• <save_path> (optional): Where to save generated audio. Defaults to results/features.
• [use-half] (optional): Add use-half at the end to enable half precision feature extraction.

For an interactive experience, launch the Gradio web interface:

python app.py

See Training.md

• • Release a more powerful foundation model covering multiple domains to provide more engaging and immersive foley creation
• • Add support for additional modalities and downstream tasks
• • Release models at different scales
• • Open-source AudioCoT dataset and automated pipeline
• • Release training scripts for ThinkSound models
• • A beginner-friendly Windows quick-start README

This project is released under the Apache 2.0 License.

Note: The code, models, and dataset are for research and educational purposes only. Commercial use is NOT permitted. For commercial licensing, please contact the authors.

📦 Third-Party Components

• Stable Audio Open VAE (by Stability AI): This repository includes a fine-tuned VAE from Stable Audio Open, licensed under the Stability AI Community License. Commercial use and redistribution require prior permission from Stability AI.

• 📘 All other code and models are released under the Apache License 2.0.

Many thanks to:

• stable-audio-tools (by Stability AI): For providing an easy-to-use framework for audio generation, as well as the VAE module and weights.
• MMAudio: For the implementation of the MM-DiT backbone in the audio domain.

If you find ThinkSound useful in your research or work, please cite our paper:

@misc{liu2025thinksoundchainofthoughtreasoningmultimodal,
 title={ThinkSound: Chain-of-Thought Reasoning in Multimodal Large Language Models for Audio Generation and Editing}, 
 author={Huadai Liu and Jialei Wang and Kaicheng Luo and Wen Wang and Qian Chen and Zhou Zhao and Wei Xue},
 year={2025},
 eprint={2506.21448},
 archivePrefix={arXiv},
 primaryClass={eess.AS},
 url={https://arxiv.org/abs/2506.21448}, 
}

✨ Feel free to open an issue or contact us via email (liuhuadai@zju.edu.cn) if you have any questions or suggestions!