VibeVoice is a novel framework designed for generating expressive, long-form, multi-speaker conversational audio, such as podcasts, from text. It addresses significant challenges in traditional Text-to-Speech (TTS) systems, particularly in scalability, speaker consistency, and natural turn-taking.
A core innovation of VibeVoice is its use of continuous speech tokenizers (Acoustic and Semantic) operating at an ultra-low frame rate of 7.5 Hz. These tokenizers efficiently preserve audio fidelity while significantly boosting computational efficiency for processing long sequences. VibeVoice employs a next-token diffusion framework, leveraging a Large Language Model (LLM) to understand textual context and dialogue flow, and a diffusion head to generate high-fidelity acoustic details.
The model can synthesize speech up to 90 minutes long with up to 4 distinct speakers, surpassing the typical 1-2 speaker limits of many prior models.
Listen to a sample of VibeVoice generating multi-speaker conversational audio:
Try it out via Demo.
| Model | Context Length | Generation Length | Weight |
|---|---|---|---|
| VibeVoice-0.5B-Streaming | - | - | On the way |
| VibeVoice-1.5B | 64K | ~90 min | HF link |
| VibeVoice-7B | 32K | ~45 min | On the way |
We recommend to use NVIDIA Deep Learning Container to manage the CUDA environment.
- Launch docker
# NVIDIA PyTorch Container 24.07 / 24.10 / 24.12 verified.
# Later versions are also compatible.
sudo docker run --privileged --net=host --ipc=host --ulimit memlock=-1:-1 --ulimit stack=-1:-1 --gpus all --rm -it nvcr.io/nvidia/pytorch:24.07-py3
## If flash attention is not included in your docker environment, you need to install it manually
## Refer to https://github.com/Dao-AILab/flash-attention for installation instructions
# pip install flash-attn --no-build-isolation- Install from github
git clone https://github.com/microsoft/VibeVoice.git
cd VibeVoice/
pip install -e .apt update && apt install ffmpeg -y # for demo
python demo/gradio_demo.py --model_path microsoft/VibeVoice-1.5B --share# We provide some LLM generated example scripts under demo/text_examples/ for demo
# 1 speaker
python demo/inference_from_file.py --model_path microsoft/VibeVoice-1.5B --txt_path demo/text_examples/1p_abs.txt --speaker_names Alice
# or more speakers
python demo/inference_from_file.py --model_path microsoft/VibeVoice-1.5B --txt_path demo/text_examples/2p_zh.txt --speaker_names Alice YunfanPotential for Deepfakes and Disinformation: High-quality synthetic speech can be misused to create convincing fake audio content for impersonation, fraud, or spreading disinformation. Users must ensure transcripts are reliable, check content accuracy, and avoid using generated content in misleading ways. Users are expected to use the generated content and to deploy the models in a lawful manner, in full compliance with all applicable laws and regulations in the relevant jurisdictions. It is best practice to disclose the use of AI when sharing AI-generated content.
English and Chinese only: Transcripts in language other than English or Chinese may result in unexpected audio outputs.
Non-Speech Audio: The model focuses solely on speech synthesis and does not handle background noise, music, or other sound effects.
Overlapping Speech: The current model does not explicitly model or generate overlapping speech segments in conversations.
We do not recommend using VibeVoice in commercial or real-world applications without further testing and development. This model is intended for research and development purposes only. Please use responsibly.