Reference
Arcane.ollama.models()
Lists the raw provider inventory envelope; it is not the application admission catalog.
This focused page is derived from the mechanically checked full member inventory.
Syntax
Arcane.ollama.models()
Parameters
None
Return value
Description
Lists the raw provider inventory envelope; it is not the application admission catalog.
Overview
Arcane.ollama.models() reads Ollama's installed tags envelope. Use it for raw
diagnostics. It is not the normalized Arcane.ai.models() inventory and must
not populate an application selector; use admitted localAI.status() models
for inference UI.
Parameters
The method takes no arguments.
Return value
It resolves to the bounded provider-native /api/tags envelope, usually with
a models array. Nested fields follow the installed Ollama version rather than
an Arcane-normalized entity contract.
Availability
This desktop Core method requires ai.models.read and is limited to Settings,
Shell, and Terminal. It is not projected to Android or browser previews.
Errors and recovery
Local provider transport, HTTP, invalid-JSON, and bounded-response failures reject as Arcane errors. Recheck local service health and retry as one owned diagnostic refresh. Do not turn a raw provider record into an authorization or verified-model claim.
Streaming, cancellation, and events
This is one non-streaming read with no method event or per-call signal. Core's provider timeout is ten seconds.
Example
const tags = await globalThis.Arcane.ollama.models();
for (const model of tags.models ?? []) {
console.info(model.name ?? model.model ?? 'Unnamed model');
}