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

Kimi K2.6 vs Kimi K3

MoonshotvsMoonshot43 shared benchmarks340 head-to-head
BenchmarkKimi K2.6Kimi K3
AA Agentic Index58.754.3
AA Intelligence45.160
AA-LCR76.782.7
AA-Omniscience6.446
aime_202696.496.7
apex_agents27.941
apexAgents28.541.3
arena_elo14611489
arena_text_factuality14551471
Artificial Analysis Coding Index61.876.2
baby_vision68.585.7
baby_vision_with_python68.585.7
browsecomp86.391.2
charxiv_rq86.784.8
coding_arena_elo15091674
critpt823.4
deepsearchqa_f192.595
Finance Agent v244.954.4
gdpval41.459.4
GDPval-AA v211901687
GPQA Diamond91.193.5
HLE37.556
hmmt_feb_202694.797
itbenchSre31.247.7
livebench_agentic_coding46.962.2
livebench_coding78.681.5
livebench_data_analysis65.178.7
livebench_instruction_following64.471.4
livebench_language75.185.5
livebench_math84.384.4
livebench_reasoning79.490.7
mathvision93.297.8
MCP Atlas68.184.2
MMMU-Pro80.183.4
OmniScience Accuracy32.847.6
OmniScience Non-Hallucination60.749.1
OSWorld-Verified73.184.8
scicode53.558.7
simpleqa_verified38.742.7
Tau 3 Banking20.633.4
TauBench V3 - Banking23.346
Terminal-Bench 2.165.988.3
toolathlon5073.2

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.