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 4 Opus vs Gemini 2.5 Flash

AnthropicvsGoogle38 shared benchmarks1917 head-to-head
BenchmarkClaude 4 OpusGemini 2.5 Flash
AA Intelligence31.717.7
AA-LCR4065.7
aider_polyglot70.761.9
AIME33.956.3
AIME 20247682.3
AIME 202575.572
AIR-Bench 202485.768.7
anthropic_red_team98.998.8
arc_agi_12733.3
ARC-AGI-28.62.5
arena_vision12071214
Artificial Analysis Coding Index3422.2
bbq99.397.7
frontiermath_tier_44.24.2
GPQA Diamond79.682.8
harmbench91.762.6
HLE12.312.1
hmmt_feb_20256064.2
IFBench53.750.3
LiveCodeBench70.476.2
livecodebench_easy99.199.1
livecodebench_hard33.148.3
livecodebench_medium80.482.5
MMMU73.779.7
scicode40.935.9
SEAL VISTA4749.1
simple_safety_tests10098
simplebench58.841.2
simpleqa22.826.9
swe_bench_bash67.628.7
SWE-bench Verified79.460.4
Terminal-Bench Hard31.113.6
vectara_answer_rate9199
vectara_avg_summary_length123.2101.5
vectara_factual_consistency8892.2
vectara_hallucination_rate127.8
xstest9798.8
τ²-Bench Telecom (AA run)73.431.6

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.