evals.report
BenchmarksLabsCompareRun guides
AnthropicClaude Sonnet

Claude Sonnet 4.6

Anthropic · Claude Sonnet. Released Feb 5, 2026.

8 results

Benchmark results 8

Compare this model
BenchmarkCategoryScoreMetricStatusDate
FrontierMathReasoning32.4%accuracyOfficialMay 30, 2026Details
DeepSWECoding31.56%% resolvedOfficialDetails
LMArenaChat preference1454source-defined ratingOfficialMay 27, 2026Details
LiveBenchReasoning75.47%scoreOfficialJan 8, 2026Details
GPQA DiamondReasoning87.4%accuracyOfficialMay 30, 2026Details
SWE-bench VerifiedCoding75.2%% resolvedOfficialMay 30, 2026Details
Humanity's Last ExamReasoning21.07%accuracyOfficialMay 31, 2026Details
MMMU-ProMultimodal75.6%accuracyOfficialApr 8, 2026Details