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 K3 vs Qwen3.6 27B

MoonshotvsAlibaba31 shared benchmarks301 head-to-head
BenchmarkKimi K3Qwen3.6 27B
AA Agentic Index54.327.5
AA Intelligence6037.7
AA-LCR82.773.3
AA-Omniscience46-20
Agents' Last Exam28.327.3
aime_202696.794.1
Artificial Analysis Coding Index76.253.7
baby_vision85.728.9
charxiv_rq84.878.4
critpt23.41.1
DeepSWE 1.16913.3
gdpval59.432
GPQA Diamond93.587.8
HLE5624
hmmt_feb_20269784.3
JobBench54.321.8
livebench_agentic_coding62.239.3
livebench_coding81.571.8
livebench_data_analysis78.770.4
livebench_instruction_following71.453.2
livebench_language85.563.3
livebench_math84.479.9
livebench_reasoning90.770.3
mathvision97.885.1
MMMU-Pro83.475.8
OmniScience Accuracy47.619.6
OmniScience Non-Hallucination49.150.7
OSWorld-Verified84.863.9
scicode58.739.8
TauBench V3 - Banking4616.7
Terminal-Bench 2.188.360.7

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.