Settings
Click Settings at the bottom of the main menu. Settings are grouped into seven sections.
General
Section titled “General”| Setting | What it does |
|---|---|
| Launch at startup | Start Owlfy when you log in (on by default) |
| Text command bar shortcut key | Option + Space / Ctrl + Space by default |
| Voice Writing key, Voice Action key | The keys for writing and for actions; they must be different |
| Mouse wheel as voice key | Use the wheel button for Voice Action (default), Voice Writing, or nothing |
| Interface language | English or Chinese |
| Appearance theme | Light, dark, or match your system |
| System permissions | Check and grant Accessibility, Screen recording, and Disk access |
- Third-party models: Owlfy’s own models are used by default. Turn this on to add models from OpenAI, Anthropic, Gemini, DeepSeek and others, or local models from Ollama and LM Studio. Each model is tested before it’s saved. See Connect a Local Model.
- AI Agent sandbox mode: All files (default) or Specified folders only. With the second option, add the folders the AI may use under Accessible folders.
- Output location: where files created by the AI are saved.
- Credits confirmation: ask before running a skill that costs more than a set number of Credits (10 by default).
- Choose and test your microphone
- Primary language: the language you usually speak
- Show live transcript
- Voice bar exit sound
- Mute system audio during voice input (on by default)
- Use speech recognition API: use your own Alibaba Cloud Bailian speech recognition key
Writing styles, output language, Smart Phrases, and the Dictionary are on the Voice Writing page in the main menu.
Clipboard
Section titled “Clipboard”Set the clipboard history shortcut and how many items to keep. The maximum depends on your membership.
Account
Section titled “Account”See your sign-in status, membership level, and renewal date, manage your devices, redeem a code, or sign out. See Account and Membership.
Privacy
Section titled “Privacy”Turn on Local Speech Recognition to turn speech into text right on your computer, without uploading recordings. The first time, Owlfy downloads the recognition model and shows its progress; it switches over automatically once ready. You can delete the model later to free up disk space.
See your version number, check for and install updates, get help and support, and follow Owlfy on social media. If a permission stops working after an update, just grant it again when prompted.
Settings elsewhere in the main menu
Section titled “Settings elsewhere in the main menu”| Page | What you can set |
|---|---|
| Voice Shortcuts | App aliases, in-app voice shortcuts, websites, folders, search engines |
| Voice Writing | Writing styles, output language, Smart Phrases, Dictionary |
| Skills | Learned skills, custom skills, MCP servers |
| Connect | Browser, Mobile Owlfy, other app connections |