- From: Christian Liebel <notifications@github.com>
- Date: Thu, 30 Oct 2025 04:33:00 -0700
- To: w3ctag/design-reviews <design-reviews@noreply.github.com>
- Cc: Subscribed <subscribed@noreply.github.com>
Received on Thursday, 30 October 2025 11:33:04 UTC
christianliebel left a comment (w3ctag/design-reviews#1093) Hi @reillyeon, the TAG is continuing to review this proposal. During the review, the following question keeps coming up: Is this API supposed to **exclusively** do on-device inference (using a model downloaded by the browser or by using a model provided by the operating system), or on an external server, or both? The proposal isn’t quite clear on this: The spec has a system to query model availability and download progress, which suggests an on-device, browser-level model, but the explainer states the API provides access to “the browser or operating system's existing language model,” and that it may allow “hybrid approaches” using external services. -- Reply to this email directly or view it on GitHub: https://github.com/w3ctag/design-reviews/issues/1093#issuecomment-3467544904 You are receiving this because you are subscribed to this thread. Message ID: <w3ctag/design-reviews/issues/1093/3467544904@github.com>
Received on Thursday, 30 October 2025 11:33:04 UTC