Privacy Policy
Sayyi · Effective 2026-09-26
Sayyi collects no data. There is no server, no account, no analytics and no advertising. Speech recognition and translation run on your iPhone, and every feature works in Airplane Mode.
What the app handles, and where it stays
- Audio from the microphone, or from what your iPhone is playing, is processed on the device. With “Record audio” on, the recording is saved in the app's own storage on the device.
- Captions and translations are produced on the device and stored with the audio.
- Settings — languages, mode, how long to keep sessions — are stored on the device.
None of this leaves your device, and we cannot access it.
Your control
- Turn off “Record audio” to keep text only.
- Set how long sessions are kept (48 hours by default) and a storage limit (10 GB by default); older sessions are deleted automatically.
- Delete any session, or all of them, at any time. Deleting the app deletes all of its data.
The only ways anything leaves the device
- Export: you hand subtitles, text, a spreadsheet or the recording to the system share sheet (AirDrop, Files, Mail…) and choose the recipient.
- Feedback: before sending, the app shows you everything that will be sent. By default that is device model, system version, memory use and the current mode — not your captions or translations, unless you choose to attach a session.
- Smart glasses (optional): if you turn on glasses in Settings, captions are sent over Bluetooth to your own glasses. They do not pass through any server.
Apple services and bundled software
Sayyi uses Apple's on-device speech recognition and translation, governed by Apple's privacy policy; the system may download a language the first time you use it. The app also contains an on-device translation model and a Bluetooth library for supported glasses. None of them collect data or connect to the internet.
Children
The app collects no data, so it collects none from children.
Changes
If a future version changes how data is handled, this page and its effective date will be updated.
Contact
allen@serenesg.com
隐私政策
说译 · 生效日期 2026-09-26
说译不收集任何数据。App 从不连接互联网：没有服务器、没有账号、没有分析工具、没有广告。开着飞行模式全部功能照常。
App 处理什么，存在哪里
- 音频：来自麦克风或本机播放的声音，只在设备上处理。开启「录音」时，音频保存在设备上 App 自己的存储区。
- 识别与译文：语音识别和翻译全部在设备上完成。结果与音频存在一起。
- 设置：模式、注释密度、保留时长、存储上限等，存在设备本地。
以上内容不会离开你的设备，我们也无法访问。
你的控制权
- 可以关闭「录音」，只保留文字记录。
- 可以设定保留时长（默认 48 小时）与存储上限（默认 10 GB），超出的旧记录自动删除。
- 可以随时手动删除单个会话或全部会话。
- 删除 App 会一并删除其全部数据。
导出与反馈：唯一会传出的路径
只有你主动操作时，内容才会离开设备，且都由你选择接收方：
- 导出：把字幕、文本、表格或录音交给系统分享面板（AirDrop、「文件」、邮件等），由你决定发给谁。
- 智能眼镜（可选）：如果你在设置里开启眼镜，字幕通过蓝牙发送到你自己的眼镜，不经过任何服务器。
- 反馈：发送前会完整展示要发出的全部内容。默认只包含机型、系统版本、内存占用、当前模式等运行信息；不包含识别到的英文和译出的中文，除非你主动勾选附带会话记录，勾选项旁会写明这一点。
苹果系统服务与内置组件
App 使用 Apple 的设备端语音识别与翻译框架。相关处理由 iOS 在设备上完成，受 Apple 自身的隐私政策约束。首次使用某种语言时，系统可能需要下载语言包。App 还内置一个在设备上运行的翻译模型，以及一个用于连接所支持眼镜的蓝牙组件；它们都不收集数据，也不联网。
儿童
App 不收集任何数据，因此也不会收集儿童的任何信息。
政策变更
如果将来版本改变了数据处理方式，我们会更新本页并修改生效日期。当前版本不进行任何数据收集。
联系
allen@serenesg.com