Reference
Arcane.localAI.status()
Returns the native host's authoritative package-policy-filtered catalog and readiness. Desktop Core includes native resource and speech admission. An admitted Android application host may instead return providerMode:"user-managed-loopback"; that status has no managed lifecycle, but it merges bounded native Kokoro/Whisper role health and catalog evidence while Android speech operations use the separate Arcane.speech bridge. Requires ai.inference.
This focused page is derived from the mechanically checked full member inventory.
Syntax
Arcane.localAI.status()
Parameters
None
Return value
Description
Returns the native host's authoritative package-policy-filtered catalog and readiness. Desktop Core includes native resource and speech admission. An admitted Android application host may instead return providerMode:"user-managed-loopback"; that status has no managed lifecycle, but it merges bounded native Kokoro/Whisper role health and catalog evidence while Android speech operations use the separate Arcane.speech bridge. Requires ai.inference.
Overview
Arcane.localAI.status() discovers currently admitted local inference and
speech choices. Use its models catalogs for application selectors and its
provider-local status fields for recovery UI. The snapshot is evidence, not a
grant; Core rechecks policy, identity, and resources at inference time.
Parameters
The method takes no arguments. Core accepts only the empty request generated by the wrapper. Concurrent callers in one Core process can share the active discovery, so reuse one application-owned refresh instead of polling from each component.
Return value
It resolves to schema-version 2 status with runtime, policy, ollama,
admission, speech, and models: {ollama, speech, transcription}. Desktop
Core reports native resource admission. Admitted Android returns
providerMode: "user-managed-loopback" and a bounded package-policy catalog;
its lifecycle remains user-managed and native desktop admission is not claimed.
Provider discovery failures normally resolve as local available,
errorCode, and empty-catalog state rather than rejecting the whole snapshot.
Availability
The method requires ai.inference. Desktop Core admits it according to package
policy. Android projects it only to explicitly admitted application identities,
using fixed host-owned loopback providers. A browser preview has no local-AI
authority.
Errors and recovery
An invalid request or package-policy/capability mismatch can reject. Otherwise, branch on the returned service-local error codes. Refresh after a documented recovery action; do not interpret a transient unavailable service as permission to switch providers or mutate models.
Streaming, cancellation, and events
Status is a non-streaming snapshot and emits no method event. The wrapper waits up to 15 seconds and exposes no signal. Core discovery is not cooperatively cancellable through this method; keep refreshes owned and avoid overlapping poll loops.
Example
const status = await globalThis.Arcane.localAI.status();
const runnableModels = status.models.ollama.filter(
function selectRunnableLocalModel(model) {
return model.runnable === true;
}
);
console.info(status.runtime.kind, runnableModels.length);