Skip to content

The Ollama sidebar is your command center for model lifecycle management. Click the 🦙 icon in the activity bar to open it.

Local Models Panel

Shows all models that have been pulled to your machine.

Model Tree

Models are grouped by family (tree view) by default. Each family node collapses/expands to show its variants (e.g., llama3.23b, 1b).

Toggle to a flat list with the Show as Flat List button (⊞) in the panel header, or back to grouped with Show as Tree (⊞).

Capability Badges

Each model item shows capability indicators:

BadgeMeaning
🧠Thinking / chain-of-thought
🛠Tool calling
👁Vision (image input)
🧩Embedding model

Status Indicators

  • Running — the model is currently loaded in Ollama's memory. Shows VRAM used and how long it has been loaded.
  • Stopped — the model is installed but not loaded.

Inline Buttons

Right side of each running or stopped model item:

ButtonWhenAction
▶ StartStoppedLoad model into memory
⏹ StopRunningUnload model from memory
🗑 DeleteStoppedRemove model from disk

Running models cannot be deleted — stop them first.

Panel Toolbar

ButtonAction
🔑Manage auth token for remote instances
🔍 FilterType to filter by model name
✕ Clear filterRemove active filter
⊞ / ⊟Toggle grouped tree / flat list
🔄 RefreshReload local model list
⊖ Collapse allCollapse all family groups

The list auto-refreshes every 30 seconds (configurable via ollama.localModelRefreshInterval).


Cloud Models Panel

Shows models from Ollama Cloud — premium hosted models that run server-side but appear in the sidebar like local models.

Authentication

Click the Login (👤) button in the panel header to authenticate. This triggers ollama login via the Ollama CLI. Once authenticated, cloud models are fetched automatically.

Inline Buttons

ButtonWhenAction
🔗AlwaysOpen the model's page on ollama.ai
▶ RunStoppedActivate the cloud model
⏹ StopRunningDeactivate the cloud model
🗑 DeleteStoppedRemove from your account

Cloud model names end with :cloud to distinguish them from local variants.


Library Panel

Browses the full Ollama model library. Models are displayed as family nodes with expandable variant children.

Pulling a Model

  1. Expand a family to see its variants (e.g., 3b, 7b, 13b, 70b)
  2. Click Pull (⬇) on the variant you want
  3. A streaming progress notification shows download progress

Variants already installed locally show a checkmark.

Filtering & Grouping

ButtonAction
🔍 FilterType to filter by model name
✕ ClearRemove active filter
⊞ / ⊟Toggle grouped / flat view
🔄 RefreshRe-fetch from ollama.ai
⊖ CollapseCollapse all families
🔗Open model page on ollama.ai

The library is fetched on startup. Use the refresh button or wait for the scheduled refresh (configurable via ollama.libraryRefreshInterval, default 6 hours).


Modelfiles Panel

See the dedicated Modelfile Manager page.

Released under the MIT License