Arcane OS Docs Development only

Reference

Arcane.ai.profile()

Gets the effective AI profile, including its canonical conversational response-length target.

  • Reference

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

Syntax

Arcane.ai.profile()

Parameters

None

Return value

Promise<AIProfile>

Description

Gets the effective AI profile, including its canonical conversational response-length target.

Overview

Arcane.ai.profile() returns the effective provider-neutral AI selection. Use it to label chat UI and to bind Arcane.ai.chat() with expectedProvider; it is not a settings mutation API.

Parameters

The method takes no arguments.

Return value

It resolves to AIProfile with provider ("ollama" or "openai"), model, configured, local, and responseLength ("low", "medium", or "high"). An OpenAI profile resolves only after Arcane proves a protected credential exists and the configured model is available to that account.

Availability

This method requires ai.inference on desktop Core. The Ollama branch is local; the OpenAI branch performs a protected account-model network request. The current Android projection does not expose this method.

Errors and recovery

For OpenAI, OPENAI_NOT_CONFIGURED, OPENAI_MODEL_REQUIRED, or OPENAI_MODEL_UNAVAILABLE requires Settings to repair the credential or model. Surface network and bounded-response failures as temporary provider failures; do not switch providers automatically.

Streaming, cancellation, and events

This is a non-streaming snapshot and emits no method event. It has no per-call signal. A profile can become stale immediately, so bind consequential chat requests with expectedProvider.

Example

const profile = await globalThis.Arcane.ai.profile();
const locationLabel = profile.local ? 'On this device' : 'Remote provider';

console.info(profile.model, locationLabel, profile.responseLength);

Reference group

AI and Ollama

Repository and reviewed source access