[AI实时语音翻译与语音生成应用] Towa Yoshida Vox Shift v1.0.44 Multilingual [WiN]（1.94GB）

P2P | Multilingual | 1.94 GB

安装方法：

– 安装预先完成的安装程序，全部激活，启动时通过设置向导选择您喜欢的用户界面语言并配置您的设置。

Vox Shift 是一款适用于支持 NVIDIA CUDA 的 PC 的 Windows 应用，可实现实时语音翻译和语音生成。它监听您选择的麦克风或音频输入设备，翻译语音，生成翻译后的语音输出，并将音频路由到语音聊天应用、游戏、OBS 和其他实时工作流程。该应用包含本地设置指南、参考语音录制、降噪、监视器输出、音板、模型准备检查和 OBS 字幕支持。

默认情况下，音频、参考语音录制、翻译文本、字幕、设置、日志和音板文件均在本地处理和存储。用户可以下载 AI 模型文件并将其存储在 PC 上。

要将翻译后的音频发送到其他应用程序，可能需要虚拟音频设备，例如 VB-Audio VB-CABLE。Microsoft Store 版本不包含 VB-CABLE 或任何第三方驱动程序安装程序。用户在安装或使用 VB-Audio 之前应查看其官方许可条款。

我们已在合作伙伴中心启用生成式 AI 声明。 Vox Shift 使用生成式 AI 技术转录用户提供的麦克风音频，进行翻译，并合成翻译后的语音/音频输出。应用商店列表中的元数据会披露此 AI 使用情况，用户可以通过应用商店列表中列出的支持联系方式报告不当的生成输出或提出其他问题。

功能包括：

– 用于实时语音工作流程的实时语音翻译
– 使用本地管理模型进行 AI 语音生成
– 录制参考语音以实现个性化语音输出
 – 监控输出以在路由前检查生成的音频
– 播放已保存的音频片段
– 降噪功能以获得更清晰的参考和实时输入
– 集成 OBS 字幕以用于流媒体工作流程
– 模型准备和就绪检查
– 用于音频路由和设备配置的设置向导
– 默认本地优先处理和存储

Vox Shift is a Windows app for real-time voice translation and voice generation on NVIDIA CUDA-capable PCs. It listens to your selected microphone or audio input device, translates speech, generates translated voice output, and helps route that audio to voice chat apps, games, OBS, and other live workflows. The app includes local setup guidance, reference voice recording, noise cancellation, monitor output, a soundboard, model preparation checks, and OBS subtitle support.

Audio, reference voice recordings, translated text, subtitles, settings, logs, and soundboard files are processed and stored locally by default. AI model files may be downloaded by the user and stored on the PC.

To send translated audio into other applications, a virtual audio device such as VB-Audio VB-CABLE may be required. The Microsoft Store version does not bundle VB-CABLE or any third-party driver installer. Users should review the official VB-Audio licensing terms before installing or using it.

We have enabled the generative AI declaration in Partner Center. Vox Shift uses generative AI to transcribe user-provided microphone audio, translate it, and synthesize translated voice/audio output. The Store listing metadata discloses this AI usage, and users can report inappropriate generated output or concerns via the support contact listed in the Store listing.

Features:

– Real-time speech translation for live voice workflows
– AI voice generation using locally managed models
– Reference voice recording for personalized voice output
– Monitor output for checking generated audio before routing
– Soundboard playback for saved audio clips
– Noise cancellation for cleaner reference and live input
– OBS subtitle integration for streaming workflows
– Model preparation and readiness checks
– Setup wizard for audio routing and device configuration
– Local-first processing and storage by default

– Install pre-done setup, all activated, choose your prefered UI language upon launch through setup wizard and configure your settings.

🏠 HomePage