Overview
Welcome to Quanty AI! Let me help you use a private, creative AI playground right in your home. This guide will walk you through the essential steps to get up and running.
To start using Quanty AI, you need two things: the Quanty AI Application and a local instance of Ollama. Download both, download a model, select your first pixel companion, and you'll be ready to chat privately with high-performance local models in minutes.
1. Install Ollama
Quanty AI uses Ollama as its LLM backend.
- Windows/Mac/Linux: Download the installer from ollama.com.
- Run the service: Once installed, ensure the Ollama service is running in your system tray.
2. Launch Quanty AI
Open the Quanty AI application. On the first launch, it will scan for your Ollama installation.
- Automatic Detection: It will find your downloaded models (like Qwen 3, Llama 3 or Mistral) automatically.
- Download a Model: If you don't have any models yet, you can download them directly from the input area, where you would type in your prompts.
3. Choose your Companion
Give your AI a face and personality with a companion. Choose a skin and give it a name, core personality and assign different roles. In these roles you can assign abilities which are mcp servers for now.
- Quanty: Our default helper. Curious and friendly.
- Pixel-Art Themed: Each companion can feature unique look and conversational styles. Everything is going to be designed by you or by me. If there is only one skin at the moment, more will follow soon.
Hi! I'm Quanty and here to assist you. Don't worry about the technical details—I'll help you navigate the setup throughout the docs so you can start using Quanty AI right away!
4. Use the Creative Studio
If you want to generate images and videos, head to Settings and enter your Runware API Key. Be aware that you use their cloud engine for generating images and videos. Your generations will be saved locally but also cached for 30 days in your Runware account. In the future, there is going to be a local integration for image generation.