Re:Apple has new Mac app coming to make running local AI mod...
by mark_l_watson on Wednesday November 11, 2026 @09:48AM
(#50454127)
Their hardware is especially good for MOE models, less suitable for dense. I use Laguna XS 2.1 6bit and Qwen-3.8-next-flash 2.6bit most because they are MOEs and run efficiently. I love Qwen-3.7:27B but as a dense model I only use it in “run something and walk away” use cases: slow.
[ Parent
| on HN ]
There are no comments on this comment yet.