Everywhere
A desktop AI assistant reads screen context on Windows and macOS, works with local models through Ollama, and uses MCP tools to carry out tasks.
Alternative to:
Everywhere is a desktop AI assistant for people who want help with the app they're working in without copying its contents into a separate chat. It reads on-screen context to answer questions and carry out tasks across apps.
- Screen context: It reads structured app information through accessibility APIs and UI automation, alongside visual screen context.
- Summaries and troubleshooting: You can summarize pages, ask what an error means and get suggested fixes, or consult financial reports, news and indicators alongside a chart.
- Desktop actions: The assistant can work with a browser and local files, dispatch sub-agents, run terminal scripts and use MCP tools; natural language system commands support service management, clearing caches and freeing occupied ports, with live shell output.
- Ways to interact: Global hotkeys, text selection, mouse shortcuts and voice interaction let you ask for help while working in another app, and responses support Markdown and mathematical formulas.
The app runs on Windows and macOS. It supports local LLMs through Ollama, custom API endpoints and cloud providers including OpenAI, Anthropic Claude, Google Gemini, DeepSeek, Moonshot Kimi, MiniMax and SiliconCloud. Local models run on your hardware; cloud models send requests to the selected external service. Everywhere also offers its own Cloud Service.
Windows integration includes Everything file search. On macOS, the assistant integrates with system apps.
