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

Gemma 4 26B A4B vs Opus 5

GooglevsAnthropic25 shared benchmarks025 head-to-head
BenchmarkGemma 4 26B A4BOpus 5
AA Agentic Index32.159.2
AA Intelligence26.163.1
AA-LCR61.778.7
AA-Omniscience-50.837.1
arena_elo14381487
arena_text_factuality14451481
Artificial Analysis Coding Index39.378
browsecomp26.390.8
coding_arena_elo13611662
critpt029.1
gdpval25.767.2
GDPval-AA v28071861
GPQA Diamond82.393.7
HLE19.364.7
HLE (with tools)17.264.7
MCP Atlas5085.8
MMMU-Pro73.884.7
OmniScience Accuracy19.160.9
OmniScience Non-Hallucination13.640.5
scicode40.355.7
SWE-bench Multilingual43.489.5
SWE-bench Pro13.879.2
SWE-bench Verified57.496
TauBench V3 - Banking1244.7
Terminal-Bench 2.13989.1

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.