All Opilot commands are accessible via the Command Palette (Ctrl+Shift+P / Cmd+Shift+P). Type "Ollama" to filter to extension commands. Most commands are also accessible through the sidebar panels and context menus.

Authentication

Ollama: Manage Ollama Auth Token

Opens a prompt to set a bearer token sent as Authorization: Bearer <token> with every Ollama API request.

Use this when your remote Ollama instance is behind an authentication proxy. Leave blank to clear the token.

Model Lifecycle

Ollama: Start Ollama Model

Starts (warms up) a local model so it is ready to respond without a cold-start delay. Also available via the ▶ button on a local model item.

Ollama: Stop Ollama Model

Unloads a running model from memory to free VRAM/RAM. Also available via the ■ button on running model items.

Ollama: Pull Ollama Model

Downloads a model from the Ollama library by name (e.g. llama3.2:3b). Opens a progress notification with live download progress. Available via the ↙ button in the panel header or the library context menu.

Ollama: Delete Ollama Model

Permanently removes an installed model and its weights from disk. Prompts for confirmation. Available via the 🗑 context menu on a local model item.

Model Parameters

Ollama: Open Model Settings

Opens the per-model settings webview panel. Also available via the ⚙ toolbar button in the Local Models view.

Use it to tune model-specific generation options:

• Temperature
• Top-P
• Top-K
• Context window (num_ctx)
• Max tokens (num_predict)
• Thinking toggle (think)
• Thinking budget (think_budget)

Changes are applied immediately and persisted per model.

Sidebar — Local Models

Ollama: Refresh Local Ollama Models

Re-fetches the list of locally installed models from the Ollama API. Triggered automatically every ollama.localModelRefreshInterval seconds.

Ollama: Filter Local Models

Opens an input box to filter the local models list by model name. The filter persists until cleared.

Ollama: Clear Local Models Filter

Removes the active filter from the local models list.

Ollama: Show as Flat List / Show as Tree (Local)

Toggles local model grouping between:

• Flat list — all models in alphabetical order
• Tree view — models grouped by model family (e.g. llama, qwen, mistral)

The current mode is reflected by the icon in the panel header.

Ollama: Collapse All Local Models

Collapses all expanded model family groups in the local models panel (tree view only).

Sidebar — Ollama Library

Ollama: Refresh Ollama Library

Re-fetches the public model list from the Ollama Library API.

Ollama: Filter Library Models

Opens an input box to filter the library model list by name or tag.

Ollama: Clear Library Models Filter

Removes the active filter from the library list.

Ollama: Show as Flat List / Show as Tree (Library)

Toggles library model view between flat list and family-grouped tree.

Ollama: Collapse All Library Models

Collapses all expanded groups in the library panel.

Ollama: Pull Ollama Model (from Library)

Downloads the selected library model tag to your local Ollama instance. Equivalent to ollama pull <model>:<tag> in the terminal.

Ollama: Open Ollama Model Page

Opens the model's page on ollama.com/library in your browser.

Sidebar — Ollama Cloud

Ollama: Login to Ollama Cloud

Opens the Ollama Cloud authentication flow in your browser. Required to access private cloud models.

Ollama: Refresh Ollama Cloud Models

Re-fetches your Ollama Cloud model catalog.

Ollama: Filter Cloud Models

Opens an input box to filter the cloud models list.

Ollama: Clear Cloud Models Filter

Removes the active filter from the cloud models list.

Ollama: Show as Flat List / Show as Tree (Cloud)

Toggles cloud model view between flat list and grouped tree.

Ollama: Collapse All Cloud Models

Collapses all expanded groups in the cloud panel.

Ollama: Run Ollama Cloud Model

Starts the selected cloud model.

Ollama: Stop Ollama Cloud Model

Stops the selected running cloud model.

Ollama: Open Ollama Cloud Model Page

Opens the model's page on the Ollama Cloud web interface.

Sidebar — Modelfiles

Ollama: New Modelfile

Opens an interactive wizard to create a new Modelfile, then opens it in the editor. See Modelfile Manager for details.

Ollama: Edit Modelfile

Opens the selected Modelfile in the VS Code editor.

Ollama: Build Model from Modelfile

Runs ollama create for the currently active or selected Modelfile. Progress is shown in a streaming VS Code notification.

Ollama: Open Modelfiles Folder

Reveals the modelfiles directory (ollama.modelfilesPath or ~/.ollama/modelfiles) in your OS file manager.

Ollama: Refresh Modelfiles

Rescans the modelfiles directory and updates the panel.

Diagnostics

Ollama: Check Server Health

Runs an immediate connectivity check against the configured ollama.host and reports whether the server is reachable. The same check is available from the status bar heartbeat item.

Ollama: Dump Performance Snapshot

Outputs a performance timing snapshot to the Opilot output channel. Useful when profiling response latency or troubleshooting provider initialization timing.

Ollama: Refresh Ollama Models

Re-fetches all model lists (local + library + cloud) in one operation.