Step-Audio-EditX

       

🔥🔥🔥 News!!

• Nov 07, 2025: ✨ Demo Page ; 🎮 HF Space Playground
• Nov 06, 2025: 👋 We release the technical report of Step-Audio-EditX.

Introduction

We are open-sourcing Step-Audio-EditX, a powerful LLM-based audio model specialized in expressive and iterative audio editing. It excels at editing emotion, speaking style, and paralinguistics, and also features robust zero-shot text-to-speech (TTS) capabilities.

📑 Open-source Plan

• Inference Code
• Online demo (Gradio)
• Step-Audio-Edit-Benchmark
• Model Checkpoints
  • Step-Audio-Tokenizer
  • Step-Audio-EditX
  • Step-Audio-EditX-Int8
• Training Code
  • SFT training
  • PPO training

Features

• Zero-Shot TTS

  • Excellent zero-shot TTS cloning for Mandarin, English, Sichuanese, and Cantonese.
  • To use a dialect, just add a [Sichuanese] or [Cantonese] tag before your text.

• Emotion and Speaking Style Editing

  • Remarkably effective iterative control over emotions and styles, supporting dozens of options for editing.
    • Emotion Editing : [ Angry, Happy, Sad, Excited, Fearful, Surprised, Disgusted, etc. ]
    • Speaking Style Editing: [ Act_coy, Older, Child, Whisper, Serious, Generous, Exaggerated, etc.]
    • Editing with more emotion and more speaking styles is on the way. Get Ready! 🚀

• Paralinguistic Editing:

  • Precise control over 10 types of paralinguistic features for more natural, human-like, and expressive synthetic audio.
  • Supporting Tags:
    • [ Breathing, Laughter, Suprise-oh, Confirmation-en, Uhm, Suprise-ah, Suprise-wa, Sigh, Question-ei, Dissatisfaction-hnn ]

For more examples, see demo page.

Model Download

Models	🤗 Hugging Face	ModelScope
Step-Audio-EditX	stepfun-ai/Step-Audio-EditX	stepfun-ai/Step-Audio-EditX
Step-Audio-Tokenizer	stepfun-ai/Step-Audio-Tokenizer	stepfun-ai/Step-Audio-Tokenizer

Model Usage

📜 Requirements

The following table shows the requirements for running Step-Audio-EditX model (batch size = 1):

Model	Setting (sample frequency)	GPU Optimal Memory
Step-Audio-EditX	41.6Hz	32 GB

• An NVIDIA GPU with CUDA support is required.
  • The model is tested on a single L40S GPU.
• Tested operating system: Linux

🔧 Dependencies and Installation

• Python >= 3.10.0 (Recommend to use Anaconda or Miniconda)
• PyTorch >= 2.4.1-cu121
• CUDA Toolkit

git clone https://github.com/stepfun-ai/Step-Audio-EditX.git
conda create -n stepaudioedit python=3.10
conda activate stepaudioedit

cd Step-Audio-EditX
pip install -r requirements.txt

git lfs install
git clone https://huggingface.co/stepfun-ai/Step-Audio-Tokenizer
git clone https://huggingface.co/stepfun-ai/Step-Audio-EditX

After downloading the models, where_you_download_dir should have the following structure:

where_you_download_dir
├── Step-Audio-Tokenizer
├── Step-Audio-EditX

Run with Docker

You can set up the environment required for running Step-Audio-EditX using the provided Dockerfile.

# build docker
docker build . -t step-audio-editx

# run docker
docker run --rm --gpus all \
    -v /your/code/path:/app \
    -v /your/model/path:/model \
    -p 7860:7860 \
    step-audio-editx

Launch Web Demo

Start a local server for online inference. Assume you have one GPU with at least 32GB memory available and have already downloaded all the models.

# Step-Audio-EditX demo
python app.py --model-path where_you_download_dir --model-source local 

Local Inference Demo

Tip

For optimal performance, keep audio under 30 seconds per inference.

# zero-shot cloning
python3 tts_infer.py \
    --model-path where_you_download_dir \
    --output-dir ./output \
    --prompt-text "your prompt text"\
    --prompt-audio your_prompt_audio_path \
    --generated-text "your target text" \
    --edit-type "clone"

# edit
python3 tts_infer.py \
    --model-path where_you_download_dir \
    --output-dir ./output \
    --prompt-text "your promt text" \
    --prompt-audio your_prompt_audio_path \
    --generated-text "" \ # for para-linguistic editing, you need to specify the generatedd text
    --edit-type "emotion" \
    --edit-info "sad" \
    --n-edit-iter 2

Technical Details

Step-Audio-EditX comprises three primary components:

• A dual-codebook audio tokenizer, which converts reference or input audio into discrete tokens.
• An audio LLM that generates dual-codebook token sequences.
• An audio decoder, which converts the dual-codebook token sequences predicted by the audio LLM back into audio waveforms using a flow matching approach.

Audio-Edit enables iterative control over emotion and speaking style across all voices, leveraging large-margin data during SFT and PPO training.

Evaluation

Comparison between Step-Audio-EditX and Closed-Source models.

• Step-Audio-EditX demonstrates superior performance over Minimax and Doubao in both zero-shot cloning and emotion control.
• Emotion editing of Step-Audio-EditX significantly improves the emotion-controlled audio outputs of all three models after just one iteration. With further iterations, their overall performance continues to improve.

Acknowledgements

Part of the code for this project comes from:

• CosyVoice
• transformers
• FunASR

Thank you to all the open-source projects for their contributions to this project!

License Agreement

• The code in this open-source repository is licensed under the Apache 2.0 License.

Citation

@misc{yan2025stepaudioeditxtechnicalreport,
      title={Step-Audio-EditX Technical Report}, 
      author={Chao Yan and Boyong Wu and Peng Yang and Pengfei Tan and Guoqiang Hu and Yuxin Zhang and Xiangyu and Zhang and Fei Tian and Xuerui Yang and Xiangyu Zhang and Daxin Jiang and Gang Yu},
      year={2025},
      eprint={2511.03601},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2511.03601}, 
}

Star History