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

DeepSeek-R1 vs Qwen2.5 Instruct 72B

DeepSeekvsAlibaba39 shared benchmarks327 head-to-head
BenchmarkDeepSeek-R1Qwen2.5 Instruct 72B
AA Intelligence18.610
AA-LCR5620.3
AA-Omniscience-31.3-52.2
aider_polyglot71.67.6
AIME 202491.423.3
AIR-Bench 202452.959
anthropic_red_team99.199.6
Artificial Analysis Coding Index24.611.9
bbq96.695.4
C-Eval91.889.2
Chinese SimpleQA (C-SimpleQA)63.752.2
CLUEWSC92.891.4
Codeforces202924.8
critpt0.60
DROP (3-shot F1)92.276.7
Fortress74.456.4
FRAMES (Acc.)8369.8
GPQA Diamond8149.1
harmbench54.672.8
HLE17.74.2
IFBench3936.9
ifeval83.387.2
LiveCodeBench84.455.5
livecodebench_pass1cot65.931.1
longbench_v258.339.4
MATH97.388.4
math_500_em9880
mmlu92.986.1
mmlu_redux93.486.8
MMLU-Pro8571.6
OmniScience Accuracy30.717.6
OmniScience Non-Hallucination10.515.3
scicode35.726.7
simple_safety_tests98.3100
simpleqa92.310.3
SWE-bench Verified57.623.8
Terminal-Bench Hard6.14.5
xstest98.897.9
τ²-Bench Telecom (AA run)11.434.5

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.