Getting Started with Xirea

Everything you need to set up and start using your private AI assistant. From download to first conversation.

Quick Start

Getting started with Xirea is straightforward. The entire process takes about 5-10 minutes, depending on your internet speed for the initial model download.

In summary: Download Xirea → Install it → Download a model → Start chatting.

That's it. No accounts, no API keys, no configuration files to edit. Xirea is designed to work out of the box so you can focus on using AI assistance rather than setting up software.

Installing on Android

1

Download the APK

Visit our GitHub Releases page and download the latest .apk file. The file will typically be named something like "Xirea-v1.0.0.apk". Make sure you are downloading from the official Danyalkhattak/Xirea repository.

2

Enable Unknown Sources

Before installing, go to your device Settings > Security (or Apps & Notifications) and enable "Install apps from unknown sources" or "Install unknown apps" for your browser/file manager app. This is required because we distribute directly, not through an app store.

3

Install the APK

Open the downloaded APK file from your Downloads folder or notification panel. Follow the installation prompts that appear. The installer will request permissions for storage access (to save models and data) — grant these for full functionality.

4

Launch and Setup

Open Xirea from your app drawer. On first launch, you will be prompted to download a language model. Select a recommended model based on your device capabilities (the app can auto-detect this) and wait for the download to complete.

Installing on Desktop

Windows

  1. Download the .exe or .msi installer from GitHub Releases
  2. Run the installer — Windows SmartScreen may show a warning since this is not a Microsoft Store app. Click "More info" then "Run anyway"
  3. Follow the setup wizard (you can choose the installation location)
  4. Launch Xirea from your Start Menu or desktop shortcut

macOS

  1. Download the .dmg file from GitHub Releases
  2. Open the .dmg and drag Xirea to your Applications folder
  3. On first launch, right-click (or Control+click) the app and select "Open" to bypass Gatekeeper
  4. Confirm you want to open the app when prompted

Linux

  1. Download the .AppImage or .deb package from GitHub Releases
  2. For AppImage: chmod +x Xirea.AppImage and run it directly
  3. For .deb: sudo dpkg -i Xirea.deb followed by sudo apt-get install -f if needed
  4. Alternatively, Flatpak and Snap builds may be available — check the releases page

Choosing a Model

Xirea supports multiple language models of varying sizes and capabilities. The choice depends on your hardware and quality requirements:

Small Models (1-2GB)

  • • Best for older devices or limited RAM
  • • Faster response times
  • • Good for simple questions and basic tasks

Medium Models (3-5GB)

  • • Recommended for most modern phones and laptops
  • • Balanced performance and quality
  • • Handles complex queries well

Large Models (7GB+)

  • • Best quality output
  • • Requires more powerful hardware
  • • Recommended for desktop use with 16GB+ RAM

The app will suggest a default model based on detected hardware, but you can always change it later in settings.

Your First Conversation

  • Start with a simple greeting or question to test the model
  • Be specific in your prompts — local models perform better with clear instructions
  • You can ask for explanations, summaries, creative writing, coding help, and more
  • Response time varies by model size and query complexity (typically 1-10 seconds)
  • Use the context window to maintain conversation history within a session

Tips for Best Results

  • Close other memory-intensive applications if responses are slow
  • Keep your device charged during long conversations (local inference uses CPU/GPU)
  • For coding tasks, specify the programming language explicitly
  • Break very long documents into smaller chunks for summarization
  • Use system prompts to set behavior (e.g., "You are a helpful tutor...")
  • Experiment with different models to find what works best for your use case

Need More Help?

If you encounter any issues or have questions not covered here, check our FAQ page or open an issue on GitHub where the community can help.

Visit FAQ →GitHub Issues →