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-V3 vs Kimi K2 (Reasoning)

DeepSeekvsMoonshot33 shared benchmarks429 head-to-head
BenchmarkDeepSeek-V3Kimi K2 (Reasoning)
AA Agentic Index1.647.9
AA Intelligence15.433.5
AA-LCR41.370.3
AA-Omniscience-37.6-21.4
aider_polyglot55.159.1
AIME 202551.394.5
Artificial Analysis Coding Index2334.8
critpt02.6
Frames73.387
gdpval024.5
GPQA Diamond68.484.5
HLE5.223.9
HMMT 202527.589.4
hmmt_feb_202529.293.3
IFBench4168.1
LiveCodeBench49.679.2
LiveCodeBench v646.983.1
longbench_v248.745.1
mmlu_redux90.594.4
MMLU-Pro81.284.6
multichallenge31.466.4
OmniScience Accuracy25.430.9
OmniScience Non-Hallucination23.325.8
scicode35.844.8
simplebench27.239.6
SWE-bench Verified4271.3
Terminal-Bench Hard15.231.1
vectara_answer_rate97.598.6
vectara_avg_summary_length81.759.2
vectara_factual_consistency93.982.1
vectara_hallucination_rate6.117.9
τ²-Bench32.574.3
τ²-Bench Telecom (AA run)47.193

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.