VECTOR WIREAI INTELLIGENCE
NVDA$1,847+3.2%MSFT$512+1.1%GOOGL$199-0.4%META$728+2.7%AMD$184-1.2%TSM$212+0.6%PLTR$98+4.1%AI IDX4,821+1.9%
PKT
SEEDRefresh Models Deals Regulatory Sources

Claude 3.5 Sonnet vs Claude Sonnet 4

AnthropicvsAnthropic34 shared benchmarks428 head-to-head
BenchmarkClaude 3.5 SonnetClaude Sonnet 4
AA Intelligence1029.8
aider_polyglot45.370.7
AIME 20241643.4
AIME 20257.474
AIR-Bench 202485.988.3
anthropic_red_team99.899.2
arena_vision11461207
Artificial Analysis Coding Index30.237.6
bbq94.997.9
CNMO 202413.160.4
Fortress1324.4
frontiermath_tier_400
GPQA Diamond67.278
harmbench98.198
HLE3.910.7
ifeval90.187.6
LiveCodeBench32.868.5
LiveCodeBench v637.248.5
math_500_em78.394
mmlu88.391.5
mmlu_redux88.993.6
MMLU-Pro7884
MMMU7274.4
MMMU-Pro54.762.4
scicode36.640
SEAL VISTA38.745.5
simple_safety_tests100100
simplebench41.445.5
simpleqa28.415.9
supergpqa48.255.7
SWE-bench Verified50.880.2
TAU-bench (airline)4660
TAU-bench (retail)69.280.5
xstest95.697.2

Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing.