Grok 4.6 Matches Claude 3.5 Intelligence at a Fraction of the API Cost

rohanpaul_ai · x · 2026-08-13

A developer analyzed the blended API costs per 1M tokens across 26 commercial LLMs. The comparison highlights that Grok 4.6 achieves 98.4% of the top model's Intelligence Index score while drastically undercutting on price.

Specifically, Grok 4.6 scores 61, just 1-point behind Claude Fable 5 Max's 62. However, Grok is 5x cheaper on input tokens ($2 vs $10/M) and over 8x cheaper on output tokens ($6 vs $50/M), presenting an unmatched intelligence-per-dollar equation.

Original post →

More from Models

Models channel →