Inference code and open weights for sound effect generative models developed at Sony AI.
Models
Model
Task
Steps
CFG
Description
Woosh-Flow
Text-to-Audio
50
4.5
Base model, best quality
Woosh-DFlow
Text-to-Audio
4
1.0
Distilled Flow, fast generation
Woosh-VFlow
Video-to-Audio
50
4.5
Base video-to-audio model
Woosh-DVFlow
Video-to-Audio
4
1.0
Distilled VFlow, fast video-to-audio
Components
Woosh-AE
— High-quality latent encoder/decoder. Provides latents for generative modeling and decodes audio from
generated latents.
Woosh-CLAP (TextConditionerA/V)
— Multimodal text-audio alignment model. Provides token latents for diffusion
model conditioning. TextConditionerA for T2A, TextConditionerV for V2A.
Woosh-Flow / Woosh-DFlow
— Original and distilled LDMs for text-to-audio generation.
Woosh-VFlow
— Multimodal LDM generating audio from video with optional text prompts.
@article{saghibakshi2025woosh,
title={Woosh: Enhancing Text-to-Audio Generation with Flow Matching and FlowMap Distillation},
author={Saghibakshi, Ali and Bakshi, Soroosh and Tagliasacchi, Antonio and Wang, Shaojie and Choi, Jongmin and
Kawakami, Kazuhiro and Gu, Yuxuan},
journal={arXiv preprint arXiv:2502.07359},
year={2025}
}
Woosh huggingface.co is an AI model on huggingface.co that provides Woosh's model effect (), which can be used instantly with this drbaph Woosh model. huggingface.co supports a free trial of the Woosh model, and also provides paid use of the Woosh. Support call Woosh model through api, including Node.js, Python, http.
Woosh huggingface.co is an online trial and call api platform, which integrates Woosh's modeling effects, including api services, and provides a free online trial of Woosh, you can try Woosh online for free by clicking the link below.
Woosh is an open source model from GitHub that offers a free installation service, and any user can find Woosh on GitHub to install. At the same time, huggingface.co provides the effect of Woosh install, users can directly use Woosh installed effect in huggingface.co for debugging and trial. It also supports api for free installation.