Library and Downloads

The Library view is your model management hub. Browse installed models, search and download new ones from HuggingFace and the Ollama registry, and manage your local model collection.

Model List

Installed models appear as cards showing the model name, parameter count, quantization level, file size, and the engine used for inference. Each card includes a quick-launch button to start chatting immediately.

Sort models by name, size, date added, or parameter count using the sort dropdown at the top of the view. Filter by quantization type, engine, or tags.

Downloading Models

Click the search icon or pressCtrl+K to open the model search panel. Type a model name, family, or tag to search across 50,000+ models on HuggingFace and the Ollama registry. Results show model name, parameter count, available quantizations, and popularity.

Download Manager

When you select a quantization and click Download, the download manager appears showing:

  • Download progress bar with percentage and ETA
  • Download speed graph (real-time throughput visualization)
  • Pause and resume buttons for each download
  • Queue management for multiple simultaneous downloads
  • Cancel download with confirmation dialog
iNote
Large models (70B+) can take significant time and disk space. OMM shows estimated download time and required disk space before starting.

Model Detail View

Click any model card to open the detail view with:

  • Model description and card from HuggingFace
  • Available quantizations and their sizes
  • Context length and architecture details
  • Tags and categories
  • Delete or re-download options

Tags

Assign custom tags to models for organization. Tags appear as colored chips on model cards and can be used as filters. Common tags includecoding, reasoning,multilingual, andvision.

Storage

Models are stored in ~/.omm/models/ by default. Change the storage location in Settings. OMM tracks total disk usage and warns when free space drops below 10 GB.

PreviousChatNextFloating Windows