Skip to content

1.5. Landing Page

After completing the first-time setup, you'll arrive at the Getting Started landing page. This is your home base in Backend.AI GO, providing quick access to essential features and an overview of your usage.

Overview

The landing page is designed to help you:

  • Track your setup progress with the interactive checklist
  • Quickly access the most-used features
  • Monitor your AI usage at a glance
  • Discover popular and recommended models
  • Get started with minimal friction

Page Layout

The landing page is organized into four main sections:

Setup Progress Checklist

Located at the top of the page, this interactive checklist guides you through the initial setup process:

Step Description
Application Installed Always complete - Backend.AI GO is running
Engine Configured Install an inference engine (llama.cpp, MLX, etc.)
First Model Downloaded Download at least one model to get started
API Server Started Start the local inference server
First Chat Completed Have your first conversation with a model

Features of the setup checklist:

  • Visual Progress Bar: Shows your overall completion percentage
  • Action Links: Click incomplete steps to navigate directly to the relevant page
  • Celebration Effect: Confetti animation when you complete all steps
  • Persistent State: Your progress is saved and remembered across sessions

Complete the Setup

Following the checklist ensures you have everything configured for the best experience. The celebration effect respects your system's reduced motion preferences.

Quick Start Section

Located on the left side, this section provides one-click access to primary features:

Action Description
Start Chatting Opens the Chat interface to begin a new conversation
Browse Models Opens the Models page to search and download models
Try Cowork Opens the Cowork interface for autonomous AI tasks

Each button shows a brief description to help new users understand the feature.

Usage Stats Section

Located on the right side, this section displays your usage statistics:

  • API Calls: Total number of inference requests made
  • Tokens Used: Combined input and output tokens processed
  • Active Conversations: Number of chat sessions

Click on this section to go to the full Statistics Dashboard for detailed analytics.

Spanning the full width at the bottom, this section showcases recommended models organized by category with enhanced cards.

Category Filters

Filter chips at the top of the section let you quickly focus on specific model types:

  • All: Show all featured models across all categories
  • General: General-purpose chat and instruction-following models
  • Reasoning: Models optimized for complex reasoning and thinking tasks
  • Vision: Vision-language models for image understanding
  • Image Generation: Text-to-image diffusion models

Click any category chip to filter the displayed models, or click "All" to see everything.

  • Popular Models: Highly-rated and frequently downloaded models
  • Recommended for Your Hardware: Models optimized for your specific system
  • New Arrivals: Recently added models to Hugging Face

Each model card displays:

  • Model Name and Description: Quick overview of what the model does
  • File Size: Storage space required for download
  • Memory Requirement: RAM/VRAM needed to run the model
  • Recommended Badge: Indicates models suitable for most systems (under 8GB)
  • System Accelerator: Shows your detected hardware (CPU, Metal, CUDA, ROCm)
  • Quick Install Button: Hover to reveal one-click download

Click on any model card to:

  • View detailed information
  • Download the model
  • See compatibility with your system

From the landing page, you can navigate to any section using:

  • Sidebar: Click icons to access Chat, Models, Agent, Engines, Statistics, etc.
  • Setup Checklist Actions: Jump to incomplete setup steps
  • Quick Start Buttons: Jump directly to primary features
  • Model Cards: Go directly to model details

Accessing the Landing Page

To return to the landing page at any time:

  • Click the Home or Getting Started icon in the sidebar
  • Use the keyboard shortcut (if configured)

Customization

The landing page adapts to your usage patterns:

  • Setup progress persists across application restarts
  • Recently used features may be highlighted
  • Model recommendations update based on your download history
  • Statistics reflect your actual usage over time

Tips for New Users

  1. Follow the Setup Checklist: Complete each step to ensure everything is configured correctly
  2. Use Quick Install: Hover over featured model cards for one-click downloads
  3. Check Recommended Models: Look for the lightning bolt badge for models that work well on your hardware
  4. Monitor Usage: Use the stats section to track your AI usage over time
  5. Explore the Sidebar: Each icon leads to a different feature area