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 Gemma 4 26B A4B

AnthropicvsGoogle23 shared benchmarks914 head-to-head
BenchmarkClaude 4 OpusGemma 4 26B A4B
AA Intelligence31.726.1
AA-LCR4061.7
AIME 202575.590
arena_vision12071238
Artificial Analysis Coding Index3439.3
GPQA Diamond79.682.3
HLE12.319.3
hmmt_feb_20256091.7
IFBench53.772.4
ifeval87.496.4
LiveCodeBench70.479.8
LiveCodeBench v647.477.1
mmlu92.977.8
mmlu_redux94.292.7
MMLU-Pro86.685.2
mmmlu88.886.3
MMMU73.778.4
scicode40.940.3
supergpqa56.561.4
SWE-bench Verified79.457.4
Terminal-Bench Hard31.125
τ²-Bench5768.2
τ²-Bench Telecom (AA run)73.443.6

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.