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-Pro vs Gemini 3 Flash Preview

DeepSeekvsGoogle40 shared benchmarks2812 head-to-head
BenchmarkDeepSeek-V4-ProGemini 3 Flash Preview
AA Agentic Index63.349.7
AA Intelligence5338.7
AA-LCR7073
AA-Omniscience-10.610.1
aime_202696.796.7
apexAgents24.327.7
arena_elo14571473
arena_text_factuality14501452
Artificial Analysis Coding Index59.442.6
coding_arena_elo15821387
critpt138.6
gdpval4935.2
GDPval-AA (Elo)15541204
GPQA Diamond90.590.4
HLE48.243.5
hmmt_feb_202695.289.4
hmmt_nov_202594.493.3
IFBench76.578
livebench73.672.4
MCP Atlas74.262
OmniScience Accuracy4353.4
OmniScience Non-Hallucination12.27.6
scicode5050.6
simplebench50.961.1
simpleqa57.968.7
simpleqa_verified57.967.4
SWE-bench Multilingual76.272.7
SWE-bench Pro55.449.6
SWE-bench Verified80.678
TauBench V3 - Banking30.120.8
Terminal-Bench 2.067.947.6
Terminal-Bench 2.172.158
Terminal-Bench Hard46.238.6
toolathlon51.849.4
vectara_answer_rate97.299.8
vectara_avg_summary_length153.890.2
vectara_factual_consistency91.486.5
vectara_hallucination_rate8.613.5
τ²-Bench Telecom (AA run)96.280.4
τ³-Bench25.817.5

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.