Arcane OS Docs Development only

Reference

Arcane.ollama.selection()

Gets Arcane's managed model preference and state.

  • Reference

This focused page is derived from the mechanically checked full member inventory.

Syntax

Arcane.ollama.selection()

Parameters

None

Return value

Promise<ArcaneModelSelection>

Description

Gets Arcane's managed model preference and state.

Overview

Arcane.ollama.selection() reads the current managed Arcane model preference and resolved runtime/provider state. Use it in Settings or Shell to explain the managed selection; use localAI.status() for an application's currently runnable model catalog.

Parameters

The method takes no arguments.

Return value

It resolves to ArcaneModelSelection: preference, recommended/effective/active variant, model and alias, default/boot settings, provider and OpenAI model, response length, credential-presence Boolean, bounded GPU diagnostics, and recommendationPending. The snapshot is diagnostic state, not admission for a later request.

Availability

This desktop Core read requires ai.models.read and is admitted only to Settings and Shell. It is not projected to Android or browser previews.

Errors and recovery

Method admission and transport errors require the correct app or host. If automatic recommendation is pending, display that state rather than treating the current model as a failed selection; refresh after the managed workflow completes.

Streaming, cancellation, and events

This is a non-streaming snapshot with no method event or signal. A concurrent selection or settings mutation can make it stale immediately.

Example

const selection = await globalThis.Arcane.ollama.selection();

console.info({
    preference: selection.preference,
    effectiveVariant: selection.effectiveVariant,
    model: selection.model,
    pending: selection.recommendationPending
});

Reference group

AI and Ollama

Repository and reviewed source access