Grok 4.7 model card: Terminal-Bench 4.0 score jumps from 20.3% to 38.0%, still trails Claude Fable 5.1

ns123abc · x · 2026-09-22

xAI's Grok 4.7 model card is out, and the Terminal-Bench 4.0 leaderboard shows a big shakeup:

A substantial gain for Grok, though it still trails the leader by roughly 20 percentage points.

Original post →

More from Models

Models channel →