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

DeepSeek-V4-Flash vs Gemma 4 E2B

DeepSeekvsGoogle19 shared benchmarks181 head-to-head
BenchmarkDeepSeek-V4-FlashGemma 4 E2B
AA Agentic Index61.37.4
AA Intelligence51.810
AA-LCR74.317
AA-Omniscience-14.3-23.6
aime_202695.837.5
Artificial Analysis Coding Index69.18.3
critpt16.60
gdpval52.90
GPQA Diamond90.843.4
HLE38.64.8
IFBench79.238
MMLU-Pro86.460
OmniScience Accuracy40.48.2
OmniScience Non-Hallucination15.667.6
scicode49.920.9
TauBench V3 - Banking39.44.5
Terminal-Bench 2.182.70.4
Terminal-Bench Hard38.63
τ²-Bench Telecom (AA run)95.622.2

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.