Beta 1.5

MistAI Desktop Assistant

Voice control, screen vision, and intelligent automation — powered by Nova, Sage, and Flux.

Voice Control Screen Vision 3 AI Personalities Live Captions
Download for Windows View Docs Get Free API Key API Docs

What is MistAI Desktop?

MistAI Desktop brings your AI assistant out of the browser and onto your system — listening for your voice, watching your screen, and taking action in real time.

Built by Kristian, it expands the MistAI ecosystem with multiple AI personalities powered by different LLM backends, each optimized for speed, reasoning, or balanced performance.

The Three Personalities

N

Nova

Fast, lightweight, creative. Powered by Gemini 2.5 Flash.

S

Sage

Strong reasoning and memory. Runs on Command-R7B.

F

Flux

Balanced speed and performance. Powered by Mistral.

Core Features

Wake Detection

Say "Mist" or "Hey Mist" to activate hands-free. Includes fuzzy matching for natural speech.

Screen Vision

Reads and clicks on UI elements using advanced OCR and computer vision.

Live Captions

Real-time overlays of what MistAI is saying and doing at the bottom of your screen.

Proactive Mode

Smart suggestions based on what you're working on — before you even ask.

App Control

Opens and focuses applications instantly with voice or text commands.

Multi-Step Actions

Chain multiple commands together for complex workflows.

Context Memory

Remembers conversations, tracks apps, and knows what you're working on.

Smart Recovery

When actions fail, MistAI consults its AI brain for recovery strategies.

Installation

Download MistAI Desktop

Available as a standalone executable for Windows, macOS, and Linux. No Python installation required — just download and run.

Download Latest Release

Windows · macOS · Linux · ~50MB · Built with PyInstaller

OCR Limitation: Bundled Tesseract OCR (screen vision) only works on Windows. macOS and Linux builds function normally but without OCR unless you install Tesseract separately.

Requirements

  • Windows 10/11 or macOS or Linux
  • Microphone — for voice control
  • Internet connection — required for AI processing
  • Vision OCR (optional) — Windows: built-in | macOS/Linux: install Tesseract
About screen vision: Windows builds include Tesseract bundled. macOS/Linux users can install it separately to enable OCR features. All other features (voice, automation, captions) work normally without it.

Caption System

The caption system provides real-time visual feedback for everything MistAI says and does. Captions appear at the bottom of your screen in non-intrusive overlays.

  • Assistant messages — green background for MistAI responses
  • System notifications — yellow background for status updates
  • Proactive suggestions — purple background for smart tips
  • Privacy first — user messages are never shown in captions
Assistant: Opening Firefox now
System: Wake word detection enabled
Suggestion: Would you like me to maximize this window?

How to Use

Wake words

Mist Hey Mist MistAI Mist AI

Example commands

"Hey Mist, open Discord"
"Mist, click on Settings"
"Search YouTube for AI tutorials"
"Scroll down"

Available Actions

open_app — Open/focus applications
click_on_text — Find and click UI elements
type_search — Type and submit text
scroll — Scroll up or down
volume — Control system volume
press_key — Press keyboard keys
maximize — Maximize current window
fullscreen — Toggle fullscreen (F11)
multi_step — Chain multiple actions

MistAI API

Integrate MistAI into your applications using a simple REST API. Send messages, select a model, and receive intelligent responses — including optional structured actions.

🔐 Authentication

All API requests require an API key. Include it in the Authorization header using the Bearer scheme.

Base URL
https://api.mistai.org/api/v1
Headers
Authorization: Bearer YOUR_API_KEY Content-Type: application/json
You can generate and manage your API key from the dashboard.

⚡ Rate Limits

Free tier includes:

  • 30 requests per minute
  • No fixed monthly cap (fair use applies)
Excessive usage may be temporarily limited to ensure system stability.

💬 POST /chat

Send a message to MistAI and receive a response from the selected model. This is the primary endpoint for interacting with the API.

Request Body
{ "message": "Explain quantum computing simply", "model": "cohere" }
Response
{ "success": true, "response": "Here's a simple explanation of quantum computing...", "model": "cohere", "timestamp": "2026-05-09T13:48:31.123456" }

🖥️ Using the API with MistAI Desktop

MistAI Desktop is powered by the same API, meaning your API key can be used seamlessly across both the web and desktop environments.

  • Use your API key to connect the desktop assistant to your account
  • All voice commands and automation requests are processed through the API
  • Switch between Nova, Sage, and Flux directly from the desktop app
The desktop app acts as a real-time interface for the API — combining voice input, screen context, and automation with AI responses.

This means you can build with the API and use the same intelligence inside MistAI Desktop without any additional setup.

Frequently asked questions

Is MistAI free to use?
Yes — MistAI Desktop is completely free and open-source. The backend API is also free to use.
Does MistAI work offline?
No, an internet connection is required for the AI backend. Computer vision and automation features work locally.
Is my data private?
MistAI only sends voice commands and screen text to the AI backend. Screenshots are processed locally and not uploaded. Conversations are not stored permanently.
Can I use MistAI on Mac or Linux?
Yes. Bundled Tesseract OCR is Windows-only, but macOS and Linux users can install Tesseract separately. Voice control, app automation, and captions all work cross-platform.
How accurate is wake word detection?
MistAI uses Google's speech recognition with fuzzy matching. It's quite accurate in quiet environments but may struggle with heavy background noise.
What's the difference between Nova, Sage, and Flux?
Nova: Fast and efficient for general tasks (Gemini 2.5 Flash)
Sage: Better at complex reasoning (Command-R7B)
Flux: Balanced performance (Mistral)
Do I need to know Python?
No — the executable requires zero coding knowledge. Just download and run.
Can I contribute to the project?
Absolutely. MistAI is open-source — check the GitHub repo, submit issues, or open a pull request.

Troubleshooting

Wake word not responding

Check microphone permissions and ensure wake word toggle is enabled in the UI.

OCR not finding text

Windows: OCR is bundled and should work automatically. macOS/Linux: Install Tesseract separately. All other features work without it.

No text-to-speech

Check system volume settings and restart MistAI if the issue persists.

Connection errors

Verify your internet connection. The status indicator in the UI shows connectivity.

Roadmap

v1.5 — Enhanced proactive suggestions with smarter context awareness
v2.0 — Cross-application workflows (e.g., "Screenshot and email it")
v2.5 — Custom wake word training
v3.0 — Plugin system for custom actions
v4.0 — Multi-monitor support with per-screen awareness

Get started

Download MistAI Desktop and bring AI to your desktop.

Download Now

Built by Kristian · MistAI Desktop © 2026