Intento is an AI-powered typing agent that lives on your desktop and helps you write smarter, faster responses wherever your cursor is.
Intento is a lightweight desktop assistant that watches your active window and provides intelligent, context-aware text suggestions in real time. It works across any application chat clients, email, code editors, browsers, and more without requiring you to switch context or copy-paste between windows.
Video.Project.1.mp4
- Cursor-aware suggestions Intento detects where you are typing and generates relevant responses on the fly.
- Multiple AI providers Supports OpenAI, Google Gemini, and local Ollama models so you can choose the backend that fits your workflow and privacy requirements.
- Custom prompt profiles Define reusable prompt templates (e.g., "Reply professionally", "Summarize this", "Translate to Spanish") and trigger them with a hotkey.
- HUD overlay A minimal heads-up display shows AI output without covering your work.
- Brain panel A persistent side panel lets you review conversation history and refine responses.
- Drag-and-drop prompt management Organize and reorder your saved prompts through a clean UI.
- PDF context Load a PDF file as background context so the AI answers with knowledge from your documents.
- Privacy-first local mode Run entirely offline using Ollama; no data leaves your machine.
Download the latest installer for Windows from the Releases page.
| Platform | Installer |
|---|---|
| Windows | Download Setup (Latest) |
Note: macOS and Linux builds are not yet available. Windows 10 / 11 (64-bit) is required.
- Download and run the installer from the link above.
- Launch Intento from the Start Menu or desktop shortcut.
- Open the Settings panel and enter your AI API key (OpenAI or Gemini) or point Intento at a running Ollama instance for fully local operation.
- Place your cursor in any text field in any app.
- Press the configured hotkey to trigger a suggestion Intento will type the response directly for you.
We recommend using the Grok API for the best experience. However, you can use other supported models if you want:
| Provider | Notes |
|---|---|
| Grok API | Recommended. Requires a Grok API key |
| Ollama (local) | Requires Ollama running locally; fully offline |
This project is source-available. See the repository for details.