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 Gemini 3.1 Pro

DeepSeekvsGoogle37 shared benchmarks235 head-to-head
BenchmarkDeepSeek-R1Gemini 3.1 Pro
AA Agentic Index3.123
AA Intelligence18.647.7
AA-LCR5679
AA-Omniscience-31.332.9
arc_agi_121.298
ARC-AGI-21.377.1
Artificial Analysis Coding Index24.668.8
browsecomp8.985.9
Chinese SimpleQA (C-SimpleQA)63.785.9
Codeforces20293052
critpt0.617.7
Fortress74.429.8
gdpval1.523.3
GPQA Diamond8194.3
HLE17.751.4
IFBench3977.1
LiveCodeBench84.491.7
mmlu92.992.6
MMLU-Pro8591
multichallenge4571.4
MultiNRC27.664.7
OmniScience Accuracy30.755.3
OmniScience Non-Hallucination10.550.1
scicode35.759
simplebench40.879.6
simpleqa_verified27.477.3
SWE-bench Multilingual30.576.9
SWE-bench Verified57.680.6
TauBench V3 - Banking6.421.4
Terminal-Bench 2.119.174
Terminal-Bench Hard6.168.5
vectara_answer_rate9799.4
vectara_avg_summary_length93.5107.7
vectara_factual_consistency88.789.6
vectara_hallucination_rate11.310.4
τ²-Bench Telecom (AA run)11.499.3
τ³-Bench6.467.1

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.