Grok Build V1.0 Lacks New Model, Only UI Fixes, Lags Behind Rivals

eyishazyer · x · 2026-08-07

Analysis reveals that despite the hype on X, xAI's Grok Build V1.0 changelog only contains UI polish and reliability fixes. The underlying model remains the same closed one running since May, with no new open-source model dropped.

Performance-wise, its only public benchmark score (70.8% on SWE-Bench Verified) is an internal number. Third-party tests show it trails Claude Opus 4.7 and GPT-5.5 by 15-18 points on the same benchmark. The actual pitch for Grok Build isn't being 'smarter' or 'open,' but rather cheaper and faster, while deeper agentic features remain locked behind the $300/mo SuperGrok Heavy tier.

Original post →

More from Companies & People

Companies & People channel →