Atlas

Compare Models

Grok 4.7 vs Kimi K3.1 Pro

Kimi K3.1 Pro leads by 3.2 capability points (≈95% interval +1.0 to +5.5): a clear lead over Grok 4.7 · P(higher) ≈ 100%

Wins 78 of 112 shared benchmarks (34 lost)

Capability vs. relative cost
Grok 4.7Kimi 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.

Grok 4.7vsKimi K3.1 Procapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins34 – 78 · 112 shared

Comparing matching setups only (66 benchmarks excluded) · suites counted once
← Grok 4.7Kimi K3.1 Pro →

Grok 4.7vsKimi K3.1 Prowins by categoryData Nov 15, 2026Data: Vals AI, SecureBio, Epoch AI and 33 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 Grok 4.7.

ProofBench v1.1Math26.091.0+65.0
FrontierMath Tier 4 (v2) - PrivateMath17.165.9+48.8
DeepSWE v1.1Code33.375.5+42.2
CloudOpsBench v1.0Agents56.519.1−37.4
LAB-Bench LitQA2 (SecureBio)Science16.953.1+36.2
BrokenArXiv 09/2026 (Native Harnesses)Math27.763.5+35.8
Furniture Assembly BenchmarkMultimodal20.855.8+35.0
BrokenArXiv 08/2026 (Native Harnesses)Math35.164.3+29.2
Terminal-Bench 4.0Code37.665.7+28.1
Legal Research Bench ImmigrationLegal45.518.2−27.3
τ³-Banking (Mercor run, bm25_grep retrieval)Agents32.659.8+27.2
LAB-Bench ProtocolQA (SecureBio)Science51.478.1+26.7
DeepsecBench — Revalidated V1 (233 issues)Code10.937.5+26.6
FrontierMath Tiers 1-3 (v2) - PrivateMath53.078.2+25.3
IOI (Vals v2)Code57.781.9+24.2
BuseyBench SVGMultimodal4.66.9+23.7
SRE Bench — Fully SolvedCode3.827.1+23.3
Terminal-Bench Science 0.1 (Vals)Science10.031.4+21.4
Legal Research Bench CriminalLegal47.428.1−19.3
ArXivMath 09/2026 (Native Harnesses)Math45.364.2+18.9
FrontierML ($200 wallet)Agents10.929.6+18.7
ArXivMath 08/2026 (Native Harnesses)Math46.564.9+18.4
FrontierMemory v1.1Agents37.855.6+17.8
Legal Research Bench Business/CommercialLegal44.227.9−16.3
ABC-Bench fragment design (SecureBio)Science100.084.2−15.8
Humanity's Last Exam DiamondGeneral QA23.639.4+15.8
Legal Research Bench Weighted PassLegal86.671.4−15.1
Legal Research Bench Civil LitigationLegal44.429.6−14.8
Bio Capabilities Index (SecureBio)Indexes141161+14.7
MBCT (SecureBio)Science46.160.7+14.6
Design Arena Overall Code CategoriesCode12701353+13.8
HoneyBench v0.2 (hack rate)Safety85.171.8+13.3
Mystery Game PuzzlesGames29.042.0+13.0
Concept Synth INDUCTION Challenge64 Round 1General QA34.446.9+12.5
GameBench 2Games72.384.0+11.7
LAB-Bench CloningScenarios (SecureBio)Science80.191.5+11.4
Real-SWE 1.1Code28.039.2+11.2
ALE-BenchCode18232132+10.9
Legal Research Bench Constitutional/Civil RightsLegal42.131.6−10.5
FrontierCode 2.0 - Pass RateCode32.322.0−10.3
Concept Synth INDUCTION Challenge100 Round 1General QA24.034.0+10.0
HPCT (SecureBio)Science53.763.5+9.8
Vals Science IndexIndexes34.844.4+9.5
DAYJOB: Finance (October 2026 rescoring)Finance & Accounting26.517.0−9.5
CWE-bench v2 Pass@1Cybersecurity54.745.3−9.4
FrontierCode 2.0 - ScoreCode28.919.9−9.0
CyberOffenseCybersecurity54.463.2+8.8
MineBench — Bradley-Terry RatingGames18481992+8.7
LMCAGeneral QA49.457.6+8.2
Creative Writing v3 EloChat & Writing20122173+8.1
Code Migration v1.1 (Vals Index v2.2 subset)Code51.859.7+7.9
AutomationBench-AA 1.0.6Agents65.673.3+7.7
APEX-Agents 1.1Professional Work68.775.9+7.2
LMArena Text Arena EloChat & Writing14421483+7.2
APEX-Accounting 1.1Finance & Accounting55.362.4+7.1
Artificial Analysis Coding Agent Index v2.1Indexes41.648.6+7.0
Visa Agentic Commerce BenchmarkAgents60.053.3−6.7
Bug Hunt BenchCode27.634.3+6.7
HLAB-Vals-1.1Legal18.311.7−6.7
OpenWorldEngineer (Mean@8)Code7.013.6+6.5
Conceptual Reasoning IndexIndexes63.167.8+6.5
DAYJOB: AccountingFinance & Accounting20.013.6−6.4
DAXBench v2Code92.686.2−6.4
BioMysteryBench (Vals Terminus 2)Science69.375.6+6.3
SimpleQA Verified (Epoch anti-abstention)General QA56.062.2+6.2
LMArena Code Arena WebDevCode16411688+6.1
WCB (SecureBio)Science57.651.5−6.1
MMMU ProMultimodal78.284.1+5.9
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal19.513.8−5.7
GDPval-AA v2.1Professional Work17181846+5.6
Tax Agent BenchFinance & Accounting69.063.9−5.1
Epoch Capabilities Index (ECI)Indexes156162+4.8
Legal Research Bench FamilyLegal18.213.6−4.5
LMArena Agent Arena Steering BurdenAgents-2.36.7+4.5
sudo L7 (October 2026 rescoring)Code12.216.7+4.5
LMArena Agent Arena Net Improvement (Steering Burden)Agents2.64.9+4.4
Access to Justice BenchmarkLegal73.769.5−4.2
VCT v2 (SecureBio)Science42.046.2+4.2
Chess Puzzles (task version 2.0)Games41.045.0+4.0
DAYJOB: LawLegal27.623.9−3.7
ACCoRDGeneral QA74.077.5+3.4
Creative Writing v3 Rubric ScoreChat & Writing15.916.6+3.4
Legal Research Bench HealthLegal76.773.3−3.3
RiemannBench (November 2026 rescoring)Math57.660.8+3.2
Longform Creative WritingChat & Writing83.079.9−3.1
Vals Finance IndexIndexes58.856.0−2.8
DTBenchGeneral QA96.093.3−2.7
LiveBench 2.0Agents74.176.7+2.6
LMArena Agent Arena Praise vs ComplaintAgents3.18.0+2.4
EnterpriseBench: CoreCraftProfessional Work68.266.1−2.1
DAYJOB: Data ScienceProfessional Work6.74.7−2.0
DualEntry Accounting AI BenchmarkFinance & Accounting88.186.1−2.0
PubMedQA (SecureBio)Health & Medicine79.281.2+2.0
OTIS Mock AIME 2024-2025Math98.1100.0+1.9
GBENCH One-shot Coding GScoreGames56.358.2+1.9
FrontierBioMLScience24.025.8+1.9
Legal Research Bench Administrative/RegulatoryLegal54.556.1+1.5
Excel Modeling Benchmark v1.1Finance & Accounting71.570.0−1.5
Vibe Code Bench v1.1Code86.284.9−1.3
LMArena Agent Arena Confirmed SuccessAgents5.12.7−1.2
AA-Briefcase v1.1 EloAgents16571678+1.2
LMArena Agent Arena Bash ReliabilityAgents9.37.2−1.1
MazeBench ASCIIGames1.02.0+1.0
RWE-benchAgents0.01.0+1.0
Finance Agent v2.1Finance & Accounting60.661.5+0.9
MysteryMechanismScience25.226.1+0.9
WMDP-Bio (SecureBio)Science88.889.7+0.9
DeepFinanceFinance & Accounting34.134.6+0.5
ABC-Bench liquid handling (SecureBio)Science98.098.3+0.3
GPQA-Bio (SecureBio)Science82.582.2−0.3
MMLU-Bio (SecureBio)Science90.991.2+0.3
LMArena Agent Arena Tool HallucinationAgents1.01.00.0

Grok 4.7vsKimi K3.1 Pronormalised score marginsData Nov 15, 2026Data: Vals AI, SecureBio, Epoch AI and 33 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…