Grok 4

In the latest public YZ Index evaluation on 2026-09-14, Grok 4 scored 79.6 overall (out of 100), ranking #2 among 11 models. The score aggregates four core dimensions — real code sandbox execution, grounding, engineering judgment, and task communication — with 100% rule-based scoring.

Grok 4 scores 81.8 on the Execution dimension, ranking #2 of 11 models. This dimension actually runs model-generated programs in an isolated sandbox to verify compilation, runtime correctness, and edge-case handling — no model-as-judge scoring.

In the WDCD instruction-decay test (2026-09-20), which measures instruction compliance under multi-turn pressure, Grok 4 scored 91.8 (out of 100), ranking #1 of 11 tested models. WDCD applies progressive distraction and social-engineering pressure, with 100% rule-based scoring.

Grok 4's API is priced at $3 per million input tokens and $15 per million output tokens (official pricing pages are verified regularly). See the Value dimension on this page for price-performance.

The YZ Index runs a full evaluation weekly and sampled evaluations daily; this page updates automatically with every public run. Current data comes from Run #323 on 2026-09-14; see the trend chart for history.