Grok 4.6 Matches GPT-5.6 Sol on Intelligence Index, Leads in Coding Benchmarks

kimmonismus · x · 2026-08-12

xAI's newly released Grok 4.6 demonstrates impressive capabilities across multiple benchmarks. According to xAI's official data, the model matches GPT-5.6 Sol at 61 on the AA Intelligence Index and outperforms it on CursorBench, FrontierCode, and AA-Briefcase.

However, Grok 4.6 still trails GPT-5.6 Sol on DeepSWE and Terminal-Bench. xAI noted that the model underwent a longer supplemental training run, with regenerated SFT trajectories and enhanced agentic reinforcement learning.

Original post →

Grok 4.6 Matches GPT-5.6 Sol on Intelligence Index, Leads in Coding Benchmarks

More from Models