Grok 4.6 Matches GPT 5.6 at 32% Lower Cost, Open Models Overpriced

downingARK · x · 2026-08-13

According to Artificial Analysis, Grok 4.6 hits a sweet spot on the Pareto frontier, matching the average score of GPT 5.6 Sol (max) while being 32% cheaper.

Contrary to popular belief, only 2 of the 16 models on the current frontier are open-source. The author notes that opting for open models to save money might actually lead to overpaying.

Related event: xAI Releases Grok 4.6: Top-Tier Performance at Unbeatable Cost(77 posts)→

Original post →

More from Models

Models channel →