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 vs Qwen2.5 Instruct 72B

AnthropicvsAlibaba33 shared benchmarks293 head-to-head
BenchmarkClaude Sonnet 4Qwen2.5 Instruct 72B
AA Intelligence29.810
AA-LCR69.720.3
AA-Omniscience0.2-52.2
aider_polyglot70.77.6
AIME 202443.423.3
AIR-Bench 202488.359
anthropic_red_team99.299.6
Artificial Analysis Coding Index37.611.9
bbq97.995.4
critpt1.10
Fortress24.456.4
GPQA Diamond7849.1
harmbench9872.8
HLE10.74.2
IFBench5536.9
ifeval87.687.2
livebench74.852.3
LiveCodeBench68.555.5
math_500_em9480
mmlu91.586.1
mmlu_redux93.686.8
MMLU-Pro8471.6
mmmlu86.574.8
MultiPL-E88.675.1
OmniScience Accuracy22.717.6
OmniScience Non-Hallucination70.915.3
scicode4026.7
simple_safety_tests100100
simpleqa15.910.3
SWE-bench Verified80.223.8
Terminal-Bench Hard31.14.5
xstest97.297.9
τ²-Bench Telecom (AA run)64.634.5

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.