evals.report
BenchmarksLabsCompareRun guides
xAIGrok

Grok 4.20 beta reasoning

xAI · Grok. Released Mar 5, 2026.

3 results

Benchmark results 3

Compare this model
BenchmarkCategoryScoreMetricStatusDate
ARC-AGI-3Reasoning0.09%accuracyOfficialMay 19, 2026Details
LMArenaChat preference1453source-defined ratingOfficialMay 27, 2026Details
LiveBenchReasoning67.96%scoreOfficialJan 8, 2026Details