Chinese models trail by ~200-250 MCI — and that gap may give them better feedback to catch up

scaling01 · x · 2026-11-22

scaling01 argues Chinese models are currently 200-250 MCI behind, and if they fall further behind they would get "slightly better feedback" — i.e., chasers have clearer improvement signals, which is part of what allows Chinese labs to catch up.

Related event: Falling behind may yield better feedback, explaining Chinese models' rapid catch-up(2 posts)→

Original post →

More from Models

Models channel →