Atlas

Compare Models

Hy4 vs Kimi K3.1 Pro

Kimi K3.1 Pro leads by 1.9 capability points (≈95% interval −0.5 to +4.3): statistically tied · P(higher) ≈ 94%

Wins 25 of 29 shared benchmarks (4 lost)

Capability vs. relative cost
Hy4Kimi K3.1 ProClosed weightsOpen weights

Each line runs through one model’s priced reasoning settings, cheapest to most expensive; grey points are other models’ priced settings in the same range. Results whose reasoning effort the source didn’t report are left out wherever a model has named effort levels. Capability and relative cost are the published estimates (relative cost is cost per task as a multiple of GPT-5 (high)); hover a setting, or use the arrow keys, to set it beside the other model’s same setting.

Hy4vsKimi K3.1 Procapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins4 – 25 · 29 shared

Comparing matching setups only (120 benchmarks excluded) · suites counted once
← Hy4Kimi K3.1 Pro →

Hy4vsKimi K3.1 Prowins by categoryData Nov 15, 2026Data: Artificial Analysis, Mercor, Neel Nanda and 3 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Kimi K3.1 Pro minus Hy4.

Terminal-Bench 4.0 (Artificial Analysis)Code37.467.2+29.8
CyberGym-E2E-AA (Artificial Analysis)Cybersecurity72.594.7+22.1
Terminal-Bench-Science 0.1Science12.934.3+21.4
Design Arena Overall Code CategoriesCode12601353+15.4
Artificial Analysis Finance & Accounting IndexIndexes38.053.3+15.3
CWE-Bench-AA (Artificial Analysis)Cybersecurity75.860.8−15.0
Artificial Analysis Legal IndexIndexes35.449.2+13.9
AA-Briefcase v1.1 EloAgents14461674+12.5
AA-Omniscience v1.1 IndexIndexes5.429.9+12.2
Artificial Analysis Intelligence Index v5.0.1Indexes34.646.5+11.9
τ³-Banking (Mercor run, bm25_grep retrieval)Agents48.159.8+11.7
Artificial Analysis Economics IndexIndexes38.349.9+11.6
DeepsecBench-AA (Artificial Analysis)Cybersecurity25.636.7+11.1
CritPt v2 (agentic, avg@5)Science59.770.6+10.9
Artificial Analysis Engineering IndexIndexes41.150.9+9.8
GDP.pdf (Artificial Analysis, All-pass)Professional Work21.631.4+9.8
APEX-Agents 1.1Professional Work85.175.9−9.2
Artificial Analysis Strategy & Ops IndexIndexes48.757.6+8.9
Artificial Analysis Healthcare & Medical IndexIndexes42.951.6+8.7
AutomationBench-AA 1.0.6Agents65.673.3+7.7
Artificial Analysis Cyber Index v1Indexes58.064.1+6.1
FrontierCode 2.0 (Artificial Analysis)Code31.425.5−6.0
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine32.838.3+5.6
NoCoT-Bench NCRI 15.2General QA94.0103+4.8
Humanity's Last Exam DiamondGeneral QA32.737.3+4.6
NoCoT-Bench NCKI (kspine v3)General QA111116+3.4
Agents' Last Exam v1.1 (ALE v1.1)Agents41.043.4+2.5
DAXBench v2Code87.286.2−1.1
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal13.613.8+0.2

Hy4vsKimi K3.1 Pronormalised score marginsData Nov 15, 2026Data: Artificial Analysis, Mercor, Neel Nanda and 3 othersVillage LLM Atlasatlas.village.dev

Win counts use each benchmark’s declared score direction. Represented suite slices and the derived Capability score are excluded from headline totals. Missing or incompatible setups are excluded, never counted as losses.
Loading Atlas data…