Skip to main content

Managing Models

Model Hub lets you discover, install, and manage AI models for your Distinct AI apps. Models are not included by default — you choose and install what works for your system.

Discovering models​

The Discover tab shows available models. Each model card displays:

  • Quality and Performance ratings to help you compare models
  • Hardware Requirements rating so you can gauge what your system needs
  • Text Rendering Supported indicator on models that can render text in images
  • A Recommended for Most tag on models suited for a wide range of hardware

Start with a model tagged Recommended for Most if you're unsure where to begin.

Choosing the right model​

Different models are designed for different purposes:

  • Some prioritize speed over quality
  • Others prioritize output quality and detail

Each model card shows Quality and Performance ratings to help you decide. If your hardware supports it, prefer models with GPU options for better performance.

Installing a model​

Selecting a model walks you through a guided install flow:

  1. Choose a Hardware Platform — select NVIDIA, Intel, or AMD
  2. Choose a Compute Device — select GPU or CPU
  3. Choose a Variant — each option shows its disk space, VRAM, and RAM requirements so you can confirm compatibility before installing
  4. Click Install

Each installation installs a single variant. You can return to install additional variants of the same model at any time.

Managing installed models​

The Inventory tab shows everything you've installed. From there you can:

  • View your installed models and variants
  • Install additional variants of a model you already have
  • Remove variants you no longer need to free up disk space

Tips​

  • Choose the model tagged Recommended for Most if you're unsure where to begin.
  • Prefer GPU variants when available for better performance.
  • Check disk space, VRAM, and RAM requirements before installing — each variant shows the exact amounts needed.
  • Install additional variants of the same model if you want to compare performance across hardware configurations.
  • Remove variants you're no longer using to free up disk space.