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 Opus 4.5 vs Gemini 2.5 Flash

AnthropicvsGoogle32 shared benchmarks274 head-to-head
BenchmarkClaude Opus 4.5Gemini 2.5 Flash
AA Agentic Index59.618.8
AA Intelligence41.917.7
AA-LCR7665.7
AA-Omniscience14-29.8
AIME 202592.872
arc_agi_18233.3
ARC-AGI-237.62.5
arena_text_factuality14581437
Artificial Analysis Coding Index47.822.2
critpt4.61.4
frontiermath_tier_44.24.2
gdpval47.311.9
GPQA Diamond8782.8
HLE30.812.1
hmmt_feb_202592.964.2
IFBench5850.3
LiveCodeBench8776.2
MMMU80.779.7
MMMU-Pro7469.1
OmniScience Accuracy46.626.1
OmniScience Non-Hallucination3924.7
scicode5035.9
SEAL VISTA46.449.1
simplebench6241.2
swe_bench_bash76.828.7
SWE-bench Verified81.560.4
Terminal-Bench Hard4713.6
vectara_answer_rate98.799
vectara_avg_summary_length114.5101.5
vectara_factual_consistency89.192.2
vectara_hallucination_rate10.97.8
τ²-Bench Telecom (AA run)89.531.6

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.