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 Sonnet 4.5 vs o3

AnthropicvsOpenAI38 shared benchmarks2116 head-to-head
BenchmarkClaude Sonnet 4.5o3
AA Agentic Index50.636.1
AA Intelligence37.431.1
AA-LCR68.373.3
AA-Omniscience-0.1-15.3
AIME 20258889.2
AIR-Bench 202489.884.5
arc_agi_163.760.8
ARC-AGI-213.66.5
Artificial Analysis Coding Index52.138.4
browsecomp24.149.7
critpt1.11.1
Fortress1716
frontiermath_tier_44.22.1
gdpval40.412.8
GPQA Diamond83.483.3
HLE19.820.3
hmmt_feb_202579.277.5
IFBench57.371.4
LiveCodeBench7184.7
longbench_v261.858.8
MathVista79.886.8
MMLU-Pro88.285
MMMU79.682.9
MMMU-Pro68.776.4
multichallenge55.360.4
MultiNRC35.845.5
OmniScience Accuracy32.938.6
OmniScience Non-Hallucination50.912.9
scicode4541
simplebench54.353.1
simpleqa_verified23.653
swe_bench_bash71.458.4
SWE-bench Verified8269.1
TAU-bench (airline)7052
TAU-bench (retail)86.273.9
Terminal-Bench Hard35.637.1
VisualToolBench6.213.7
τ²-Bench Telecom (AA run)78.180.7

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.