⌘R长按Hold
说话Talk
按住录音，松开即转写、润色（去口水词、补标点、修正技术词）并填入输入框Hold to record; release to transcribe, polish (filler words, punctuation, technical terms), and insert
对着 AI 编码工具说话，出来就是干净的指令 · 免费Talk to your AI coding tools — get clean prompts · Free
长按说话
长按说话
松开出字
单击发送
双击取消
Hold to talk
Release to type
Click to send
Double-click to cancel
一个键，三个动作One key, three actions
说话、发送、取消Talk, send, cancel
默认 macOS 用右 Command、Windows 用 F13，都可以在设置里更换Right Command on macOS, F13 on Windows — change it in Settings
⌘R单击Click
发送Send
模拟 Return 发送当前内容，可在设置中关闭Simulates Return to send the current text; disable it in Settings
⌘R双击Double-click
取消Cancel
录音时丢弃；文字刚填入且焦点未变化时撤销Discard a recording, or undo inserted text before focus changes
隐私与处理方式Privacy and processing
默认本地，云端可选Local by default, cloud optional
识别和润色可分别选择本地或云端；本地未就绪时只会提示。Choose local or cloud separately for speech and polishing. If local processing is unavailable, Anelo tells you.
- 本地：Local: 音频和文字留在本机audio and text stay on your computer
- 云端：Cloud: 仅在你主动选择并配置 Key 后启用enabled only after you choose it and add a key
- Key：Keys: 交给系统加密保管（macOS 钥匙串 / Windows DPAPI）encrypted by the OS (macOS Keychain / Windows DPAPI)
完全本地Fully local
离线使用，隐私优先Private and offline
Whisper 识别，Ollama 润色。首次需安装组件和模型。Whisper handles speech and Ollama handles polishing. Initial setup is required.
云端服务Cloud services
使用自己的 KeyUse your own key
支持 Groq、OpenAI 和 Gemini。Groq 有免费方案，通常足够个人日常使用；具体限额以 Groq 官方为准。Supports Groq, OpenAI, and Gemini. Groq's free plan is usually enough for everyday personal use; limits are set by Groq.
开始使用Get started
装好就能用Get set up
macOS 三步，Windows 两步。Three steps on macOS, two on Windows.
macOS 13+ · Apple 芯片与 Intel · 已通过 Apple 公证 · Windows 10 1809+ · x64 与 ARM64 macOS 13+ · Apple silicon & Intel · Apple-notarized · Windows 10 1809+ · x64 & ARM64
macOS
- 拖入 Applications — 打开 DMG，把 Anelo 拖进「应用程序」并启动 Drag to Applications — open the DMG, drag Anelo in, and launch it
- 完成权限 — 允许麦克风、辅助功能、输入监控三项，然后重开一次 Grant access — allow Microphone, Accessibility and Input Monitoring, then reopen once
- 选择处理方式 — 云端填入自己的 Key；完全离线则安装 Whisper 和 Ollama Choose processing — add your cloud key, or install Whisper and Ollama for offline use
Windows
无需系统授权No permissions needed- 运行安装包 — 装完自动启动，托盘上出现 Anelo 图标 Run the installer — it launches on finish and appears in the system tray
- 选择处理方式 — 云端填入自己的 Key；完全离线则安装 Whisper 和 Ollama Choose processing — add your cloud key, or install Whisper and Ollama for offline use
更自在的操控方式A more comfortable way to work
摆脱键盘，手机也能远程操控Control Anelo from your phone, away from the keyboard
同一可信 Wi-Fi 下扫码打开遥控页，无需安装 App。手握手机即可说话、发送和取消，不必一直把手放在键盘上；录音和处理仍在本机。On the same trusted Wi-Fi, scan to open the remote with no app to install. Hold your phone to talk, send, or cancel without keeping your hands on the keyboard; recording and processing stay on your computer.
iPhone 与 Android 手机均可用Works on iPhone and Android
就绪Ready
按住说话Hold to talk
点一下发送 · 双击取消Tap to send · Double-tap to cancel
常见问题FAQ
几个你可能会先问的A few things people ask first
macOS 为什么需要三项系统权限？Why does macOS need three system permissions?
麦克风负责录音；输入监控检测触发键；辅助功能识别输入框并执行填入、发送和取消。Windows 不需要任何系统授权。Microphone records, Input Monitoring detects the trigger key, and Accessibility inserts, sends, and cancels in the current field. Windows needs no system permissions at all.
本地模式断网后还能使用吗？Does local mode work without an internet connection?
可以。安装好 Whisper、Ollama 和模型后，识别与润色均可离线运行。Yes. After installing Whisper, Ollama, and their models, speech and polishing run offline.
一定需要 API Key 吗？Is an API key required?
本地模式不需要；只有选择云端识别或润色时才需要。Not for local mode. You need one only for cloud speech or polishing.
手机遥控会使用手机麦克风吗？Does the phone remote use the phone microphone?
不会。手机只发送按键动作，录音、转写和润色都在本机。No. The phone sends controls; recording, transcription, and polishing stay on your computer.
所有输入框都能使用吗？Does it work in every text field?
支持大多数标准输入框、终端和开发工具；少数自定义编辑器可能不同。It works in most standard fields, terminals, and developer tools. Some custom editors may differ.