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 Gemini 3 Flash Preview

AnthropicvsGoogle38 shared benchmarks731 head-to-head
BenchmarkClaude Sonnet 4Gemini 3 Flash Preview
AA Agentic Index39.249.7
AA Intelligence29.838.7
AA-LCR69.773
AA-Omniscience0.210.1
AIME 20257499.7
arc_agi_163.784.7
ARC-AGI-213.633.6
arena_vision12071260
Artificial Analysis Coding Index37.642.6
critpt1.18.6
frontiermath_tier_404.2
gdpval31.235.2
GPQA Diamond7890.4
harmbench9870
HLE10.743.5
IFBench5578
livebench74.872.4
mmmlu86.591.8
MMMU-Pro62.481.2
OmniScience Accuracy22.753.4
OmniScience Non-Hallucination70.97.6
OSWorld-Verified42.265.1
scicode4050.6
simplebench45.561.1
simpleqa15.968.7
swe_bench_bash64.975.8
SWE-bench Multilingual56.972.7
SWE-bench Pro42.749.6
SWE-bench Verified80.278
TauBench V3 - Banking16.720.8
Terminal-Bench 2.136.358
Terminal-Bench Hard31.138.6
vectara_answer_rate98.699.8
vectara_avg_summary_length145.890.2
vectara_factual_consistency89.786.5
vectara_hallucination_rate10.313.5
τ²-Bench Telecom (AA run)64.680.4
τ³-Bench13.817.5

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.