以声为笔，无需键盘。让不便于打字的人，也能自由表达。按住右 Cmd 说话，松手即输入。
Your voice is your keyboard. Built for people who can't easily type. Hold right Cmd, speak, release — words appear at your cursor.
对有运动障碍、手指不便或无法正常打字的用户，按住右 Cmd 说话远比逐键打字轻松。松手自动粘贴到光标位置——无需复制、切换窗口。
For those with motor impairments, arthritis, or limited finger mobility, holding right Cmd to speak is fundamentally easier than pressing individual keys. Release to paste — no copying, no window switching.
所有识别在本机完成。不需要网络、不需要云服务、不需要账号。
All recognition runs locally. No internet, no cloud, no account required.
说话时文字实时出现在浮动窗口中——轻量模型流式识别，LLM 模型松手校正。
Text appears in a floating window as you speak — lightweight model streams in real time, LLM corrects on release.
支持中文和英文混合识别。可选标准与增强校正模型。
Supports Chinese and English mixed recognition. Standard or Enhanced correction models.
备忘录、微信、浏览器、终端——光标在哪，文字就到哪。
Notes, Messages, browser, terminal — wherever your cursor is, your words go there.
CPU 占用约 5–10%，内存约 160MB。Mac 上同时做其他事完全不受影响。
~5–10% CPU, ~160MB memory. Runs quietly while you do everything else.
macOS 14.0 或更高版本，需配备 Apple Silicon (M1/M2/M3/M4)。
macOS 14.0 or later. Requires Apple Silicon (M1/M2/M3/M4).
App Store 也可下载。含 14 天免费试用。
Also on the App Store. 14-day free trial included.