1 article found
A developer runs LFM2.5-2.6B on a OnePlus 13 at 17 tok/s with pure CPU inference. This is what happens when edge AI stops being a demo and becomes production.