VoiceGen Support
Offline voiceovers in English and Chinese, generated on your Mac.
Contact
Email guoduzhe@gmail.com. Please include your macOS version, your Mac model and what you were doing when the problem happened.
Requirements
- A Mac with Apple silicon and macOS 14 or later.
- About 2 GB of free space for the app, which includes the Standard voice model.
- The optional High Quality model needs about 9 GB of space and at least 16 GB of memory.
Common questions
Does VoiceGen need the internet?
No. Voiceovers are generated entirely on your Mac. VoiceGen only goes online when you download a model, or when you open the Models page to check for model updates.
How do I add pauses or switch voices?
Write [pause 0.8] for a 0.8-second pause, or [voice Vivian] to switch the voice for the following sentences. You can also use the “+ Pause” and “+ Voice” buttons above the script.
A sentence sounds wrong. What can I do?
Click the regenerate button on that sentence. VoiceGen creates a new take and keeps the old ones, so you can switch back at any time. Editing the script only regenerates the sentences you changed.
How do I use VoiceGen from Claude, Codex or Cursor?
Open Connect AI, turn on Allow AI apps to use VoiceGen, then copy the configuration for your app. Your AI app can then ask VoiceGen to generate a voiceover. Files are saved to Downloads/VoiceGen. If macOS asks whether your AI app may access the Downloads folder, click Allow.
Can I use VoiceGen with Shortcuts or Siri?
Yes. Use the “Generate Voiceover” action in Shortcuts, or say “Generate a voiceover with VoiceGen”.
Can I clone a real person’s voice?
No. VoiceGen offers built-in voices, and with the High Quality model you can design a new voice from a text description. It does not clone voices from recordings of real people.
Where are my projects stored?
On your Mac, inside VoiceGen’s app data. Use Export in the Studio to save a WAV file and SRT subtitles to any folder you choose.
VoiceGen 技术支持
在你的 Mac 上离线生成中英文配音。
联系方式
邮箱 guoduzhe@gmail.com。来信请附上 macOS 版本、Mac 型号，以及出问题时你在做什么。
系统要求
- Apple 芯片的 Mac，macOS 14 或更高版本。
- 约 2 GB 可用空间（含标准语音模型）。
- 可选的高质量模型需要约 9 GB 空间和至少 16 GB 内存。
常见问题
需要联网吗？
不需要。配音完全在本机生成。只有下载模型，或打开“模型”页检查更新时才会联网。
怎么加停顿、换音色？
在文案里写 [pause 0.8] 表示停顿 0.8 秒，写 [voice Vivian] 表示后面的句子换成 Vivian。也可以点文案上方的“+ 停顿”和“+ 音色”。
某一句读得不好怎么办？
点这一句右侧的重新生成按钮。新版本会保留旧版本，随时可以切回去。修改文案后，也只会重新生成改过的句子。
怎么在 Claude、Codex、Cursor 里用？
打开“连接 AI”，打开“允许 AI 应用使用 VoiceGen”，再复制对应应用的配置。之后你的 AI 应用就能请 VoiceGen 生成配音，文件保存在“下载/VoiceGen”。如果系统询问是否允许该应用访问“下载”文件夹，点“允许”。
能用快捷指令或 Siri 吗？
可以。在快捷指令里使用“生成配音”动作，或者对 Siri 说“用 VoiceGen 生成配音”。
能克隆真人的声音吗？
不能。VoiceGen 提供内置音色；使用高质量模型时，还可以用文字描述设计新音色。它不会根据真人录音克隆声音。
项目存在哪里？
存在本机的 VoiceGen 数据目录里。在工作台点“导出”，可以把 WAV 和 SRT 字幕存到任意文件夹。