Atlas

Compare Models

Grok 4.6 vs Grok 4.7

Grok 4.7 leads by 0.6 capability points (≈95% interval −1.4 to +2.7): statistically tied · P(higher) ≈ 72%

Grok 4.6 wins 111 of 238 shared benchmarks (106 lost, 21 tied)

Capability vs. relative cost
Grok 4.6Grok 4.7Closed 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 4.6vsGrok 4.7capability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins111 – 106 · 238 shared, 21 tied

Comparing matching setups only (4 benchmarks excluded) · suites counted once
← Grok 4.6Grok 4.7 →

Grok 4.6vsGrok 4.7wins by categoryData Nov 15, 2026Data: Vals AI, Artificial Analysis, Mercor and 57 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Grok 4.7 minus Grok 4.6.

Design Creativity BenchCode43.462.2+59.8
Harvey's Legal Agent Benchmark - Capital MarketsLegal0.040.0+40.0
LLM Creative Story WritingChat & Writing-2.90.4+37.4
Terminal-Bench 2.1 Extended (Mercor held-out)Code45.511.8−33.7
AA-PersonalityBench Slop indexChat & Writing35.38.6−26.9
Toloka Arena (v3)Agents56.031.0−25.0
ProofBench v1.1Math51.026.0−25.0
SWE-bench Verified Extended (Mercor held-out)Code78.055.3−22.7
FoodTruck BenchGames5953742561−21.8
Harvey's Legal Agent Benchmark - Banking FinanceLegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Corporate GovernanceLegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Data Privacy CybersecurityLegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Energy Natural ResourcesLegal40.020.0−20.0
Harvey's Legal Agent Benchmark - Intellectual PropertyLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Structured Finance SecuritizationLegal20.00.0−20.0
Furniture Assembly BenchmarkMultimodal40.020.8−19.2
BrowseComp Extended (Mercor held-out)Search21.539.5+18.0
KernelBench Mega Normalized (RTX PRO 6000)Code0.018.0+18.0
ChessBench CoherenceGames69.286.3+17.1
CloudOpsBench v1.0Agents39.756.5+16.8
BioSecBench-Refusal V1 Clean RerunsSafety45.662.4+16.8
LAB-Bench ProtocolQA (SecureBio)Science68.151.4−16.7
CadQueryEval (Corrected Geometry)Code72.088.0+16.0
LLM Chess EloGames443822+15.6
AA-PersonalityBench Organization indexChat & Writing14.329.6+15.4
AA-PersonalityBench Imagination indexChat & Writing20.035.3+15.4
FrontierMath Tier 4 (v2) - PrivateMath31.717.1−14.6
LAB-Bench CloningScenarios (SecureBio)Science94.280.1−14.1
AA-AnalystAgent (pass^5)Professional Work41.355.0+13.8
SWE-Marathon v1.1Code31.945.6+13.8
AA-PersonalityBench Warmth indexChat & Writing31.544.8+13.5
AA-PersonalityBench Transparency indexChat & Writing71.458.1−13.5
τ³-Banking (Mercor run, bm25_grep retrieval)Agents45.732.6−13.1
FrontierMath Tiers 1-3 (v2) - PrivateMath66.053.0−13.0
SnorkelUnderwrite 2.0Finance & Accounting17.830.5+12.7
ComplexConstraintsChat & Writing36.548.9+12.4
Terminal-Bench 4.0Code25.337.6+12.3
LAB-Bench LitQA2 (SecureBio)Science28.916.9−12.0
SAGEMath28.940.8+11.9
AA-PersonalityBench Informativeness indexChat & Writing44.833.4−11.5
AA-PersonalityBench Expressiveness indexChat & Writing63.875.3+11.5
ChessBench EloGames8911076+11.0
CWE-bench v1 Programmatic Pass@1Cybersecurity57.068.0+11.0
Veracode GenAI Code Security (Fall 2026)Cybersecurity58.069.0+11.0
EEBench - release-reported electrical engineering scoreScience53.064.0+11.0
CritPt v2 (agentic, avg@5)Science62.351.4−10.9
Toloka Arena (v1)Agents49.739.3−10.4
IOI (Vals v2)Code47.657.7+10.1
Vibe Code Bench v1.1Code76.286.2+9.9
Olam Social Poker EloGames14991508+9.8
AA-PersonalityBench Conviction indexChat & Writing92.482.9−9.6
LMArena Agent Arena Net ImprovementAgents1.14.0+9.5
Artificial Analysis Coding Agent Index v1.5Indexes47.056.3+9.3
HardGeoBench Avg DistanceMultimodal18072104−9.2
Vending-Bench 2Games904710537+9.0
CAD ArenaAgents34.825.8−9.0
FrontierCode 2.0 (Artificial Analysis)Code19.828.7+8.9
APEX-Agents 1.1Professional Work77.368.7−8.6
SWE-bench MultilingualCode92.784.2−8.5
BrowseComp (Mercor 130-question subset)Search84.075.8−8.2
Chess Puzzles (task version 2.0)Games33.041.0+8.0
SnorkelLegalLegal40.732.8−7.9
HPCT (SecureBio)Science61.653.7−7.9
KernelBench CUDA Normalized (RTX PRO 6000)Code37.745.2+7.5
WCB (SecureBio)Science50.357.6+7.3
SimpleQA Verified (Epoch anti-abstention)General QA49.356.0+6.7
MetagenomicsBench V0Science42.048.7+6.7
MetagenomicsBench V1Science42.048.7+6.7
ChessBench AccuracyGames67.874.0+6.6
RuneBenchGames5.85.4−6.5
VariantBench V1 Clean RerunsScience39.045.5+6.5
DAXBench v2Code86.292.6+6.4
MedXpertQA MM (200-question set)Health & Medicine75.269.0−6.2
Eval Connections One-Shot with TrapsGames91.085.0−6.0
DeepSWE v1.1Code65.271.0+5.8
ChartographyMultimodal20.514.7−5.8
ARC-AGI-2General QA67.161.4−5.7
GDP.pdfProfessional Work17.222.8+5.6
DAYJOB: HealthcareHealth & Medicine2.88.4+5.6
SpeciEval (Refusal-Aware Retries, Historical)Safety90.896.4+5.6
SpeciEval (Pooled English Runs)Safety90.896.4+5.6
SWE-bench VerifiedCode84.679.1−5.5
MysteryMechanismScience30.625.2−5.4
AA-Briefcase v1.1 EloAgents15401636+5.3
Vals RSI Index v1.1 (four tasks)Agents25.520.2−5.3
DeepsecBench — Revalidated V1Code16.210.9−5.2
DeepsecBench — Revalidated V1 (233 issues)Code16.110.9−5.2
LMArena Agent Arena Net Improvement (Steering Burden)Agents-0.22.6+5.2
Mystery Game PuzzlesGames34.029.0−5.0
APEX-Accounting 1.1Finance & Accounting60.355.3−5.0
GeneBench-Pro Public Release SubsetScience15.010.0−5.0
WellnessBenchSafety43.048.0+5.0
HardGeoBench <=100miMultimodal32.527.5−5.0
GPQA-Bio (SecureBio)Science87.482.5−4.9
SnorkelFinance 2.0Finance & Accounting25.120.2−4.9
CursorBench 4.0Code41.446.3+4.9
MedCodeHealth & Medicine44.749.6+4.8
ScBench-Long V1 Clean RerunsScience15.911.4−4.5
Clipping Bench v1.0Games805384−4.5
SnorkelRevOpsProfessional Work15.811.3−4.5
SWE-TogetherCode60.665.1+4.5
Artificial Analysis Intelligence Index v5.0.1Indexes31.035.5+4.5
Roboflow Vision Evals (October 2026, five tasks)Multimodal59.163.5+4.4
APEX-SWECode57.753.4−4.3
SnorkelManufacturingProfessional Work16.412.1−4.3
MBCT (SecureBio)Science50.446.1−4.3
Excel Modeling BenchmarkFinance & Accounting62.767.0+4.3
FrontierSWE V2Code25.329.5+4.2
SWE Atlas Codebase QnA Extended (Mercor held-out)Code19.315.3−4.0
Arena Image-to-WebDevCode16071589−4.0
LMArena Agent Arena Confirmed SuccessAgents-2.65.1+3.8
Vision Arena OverallMultimodal12631249−3.7
GameBench 2Games68.772.3+3.6
MedXpertQA MM Extended (Mercor held-out)Health & Medicine48.452.0+3.6
Terminal-Bench 4.0 (Artificial Analysis)Code21.224.7+3.5
SpatialBench V1 Clean RerunsScience66.169.6+3.5
YC-BenchGames15682411666389+3.4
BioSecBench-Function V1 Clean RerunsScience39.943.3+3.4
Terminal-Bench 2.1Code88.485.0−3.4
Terminal-Bench-Science 0.1Science6.29.5+3.3
SWE Atlas - Codebase QnACode50.347.0−3.3
Vals Science IndexIndexes38.134.8−3.3
Extended NYT Connections 940 (Quadratic v1)Games80.076.8−3.2
Roboflow Vision Evals (September 2026 protocol)Multimodal68.771.9+3.2
CVDP (517-problem set)Code58.054.9−3.1
Bio Capabilities Index (SecureBio)Indexes137141+3.0
LMArena Code Arena WebDevCode16181641+3.0
AA-LCR Extended (Mercor held-out)Search65.268.2+3.0
Design Arena Overall Code CategoriesCode12881270−3.0
SpatialBench-Long V1 Clean RerunsScience27.424.4−3.0
BioMysteryBench (Vals Terminus 2)Science72.269.3−3.0
AutoResearchExam (hidden-test AUARC)Agents0.50.5+3.0
Vals Index v2.1Indexes52.154.9+2.9
MedScribeHealth & Medicine86.589.4+2.8
Medical Long Context Reasoning (MLCR-AA)Health & Medicine12.215.0+2.8
LMArena Agent Arena Praise vs ComplaintAgents-2.53.1+2.8
SM BenchSafety89.086.3−2.7
ARC-AGI-1 Semi-PrivateGeneral QA87.590.2+2.7
EnterpriseBench: CoreCraftProfessional Work65.668.2+2.6
GDPval-AA v2.1Professional Work16581717+2.6
LMArena Text Arena EloChat & Writing14561442−2.5
HardGeoBench <=10miMultimodal20.017.5−2.5
ACCoRDGeneral QA71.674.0+2.4
Tax Agent BenchFinance & Accounting71.469.0−2.4
MMMU ProMultimodal80.578.2−2.3
Access to Justice BenchmarkLegal71.373.7+2.3
TxBench Preclinical Pharmacology V1 Clean RerunsScience57.054.7−2.3
VIALSMultimodal20.518.2−2.3
APEX-1 (Mercor, mean score)Professional Work67.765.4−2.3
MineBench — Bradley-Terry RatingGames18111848+2.2
LLM Debate BenchmarkChat & Writing16041620+2.2
Artificial Analysis Intelligence Index v4.3.2Indexes44.346.4+2.1
Humanity's Last Exam DiamondGeneral QA21.523.6+2.1
Immigration BenchLegal66.068.0+2.1
ReactBench (39 tasks)Code33.831.8−2.1
Harvey HLAB — Criteria Pass RateLegal92.590.5−2.0
Arena Alignment IndexSafety80.882.7+2.0
FrontierCode 1.1 Extended - ScoreCode61.359.4−2.0
LegalBenchLegal86.384.4−1.9
Snorkel WorkplaceAgentsProfessional Work17.115.2−1.9
FrontierCode 1.1 Extended - Pass RateCode67.065.2−1.8
CyberBench v1.1 OverallCybersecurity67.769.5+1.8
CritPt (avg@5)Science19.718.0−1.7
RSI-Exam 0.1 (Full 88)Agents0.40.4+1.7
Snorkel Agentic Coding 2.0Code33.631.9−1.7
DAYJOB: FinanceFinance & Accounting12.814.5+1.7
Concept Synth INDUCTION Challenge64 Round 1General QA35.934.4−1.6
Epoch Capabilities Index (ECI)Indexes158156−1.5
AutomationBench-AA 1.0.6Agents67.065.6−1.4
Finance Agent v2Finance & Accounting53.752.3−1.4
Terminal-Bench Science 0.1 (Vals)Science11.410.0−1.4
Biopharma Bench V0.1Health & Medicine63.064.4+1.4
LMArena Agent Arena SteerabilityAgents3.30.5−1.4
TxBench Oligo Discovery V1 Clean RerunsScience41.740.3−1.4
DTBenchGeneral QA97.396.0−1.3
SciCode 1.0.1 (Artificial Analysis)Code56.557.8+1.3
GPQA DiamondGeneral QA94.092.7−1.3
Snorkel Digital Work IndexIndexes20.519.3−1.2
Public Benefits Bench v1.1Professional Work66.865.6−1.2
OTIS Mock AIME 2024-2025Math99.298.1−1.1
Senior SWE-BenchCode26.327.4+1.1
CheatBenchSafety79.178.0+1.1
Vals Index v2Indexes59.260.2+1.0
AA-LCR v1.1Search81.080.0−1.0
Concept Synth INDUCTION Challenge100 Round 1General QA25.024.0−1.0
RedlineBenchLegal48.549.5+1.0
Legal Research BenchLegal48.147.1−1.0
LMCAGeneral QA48.549.4+0.9
GBENCH One-shot Coding GScoreGames55.456.3+0.9
CharXiv (descriptive and reasoning combined)Multimodal92.992.0−0.9
CharXiv Extended (Mercor held-out)Multimodal77.378.2+0.9
MMLU-Bio (SecureBio)Science91.890.9−0.9
ScBench V1 Clean RerunsScience57.658.5+0.9
LiveBench 2.0Agents74.974.1−0.8
Conceptual Reasoning IndexIndexes62.563.1+0.8
AA-Omniscience IndexIndexes30.532.0+0.8
LMArena Agent Arena Bash RecoveryAgents5.77.2+0.8
TxBench Antibody Discovery V1 Clean RerunsScience43.542.8−0.7
AdvancedIFChat & Writing80.980.2−0.7
Blueprint-Bench 2Multimodal0.30.3−0.7
ObviousBenchGeneral QA100.099.3−0.7
MMLU ProGeneral QA89.488.8−0.6
ProgramBenchCode1.00.5−0.5
Artificial Analysis Progress Index v1Indexes322323+0.4
Snorkel SWE-bench CLICode5.76.1+0.4
BuseyBench SVGMultimodal4.64.6−0.4
LMArena Agent Arena Bash ReliabilityAgents10.09.3−0.4
EpiBench V1 Clean RerunsScience21.120.8−0.3
Bug Hunt BenchCode27.327.6+0.3
ForecastBench Preliminary Dataset Brier IndexForecasting61.060.7−0.3
ARC-AGI-3 (Semi-Private)General QA2.11.8−0.3
LMArena Agent Arena Steering BurdenAgents-1.8-2.3−0.3
Code MigrationCode44.644.8+0.2
Humanity's Last Exam (Text-Only)General QA42.943.1+0.2
Realm Financial ReasoningFinance & Accounting44.644.5−0.1
ALE-BenchCode182218230.0
LMArena Agent Arena Tool HallucinationAgents1.01.00.0
AIME 2025Math100.0100.00.0
BioSecBench-Surveillance V1 Clean RerunsScience48.048.00.0
HardGeoBench Country AccuracyMultimodal43.843.80.0
Harvey's Legal Agent Benchmark - Antitrust CompetitionLegal20.020.00.0
Harvey's Legal Agent Benchmark - Arbitration International Dispute ResolutionLegal0.00.00.0
Harvey's Legal Agent Benchmark - Bankruptcy RestructuringLegal0.00.00.0
Harvey's Legal Agent Benchmark - Corporate M&ALegal20.020.00.0
Harvey's Legal Agent Benchmark - Emerging Companies Venture CapitalLegal0.00.00.0
Harvey's Legal Agent Benchmark - Employment LaborLegal0.00.00.0
Harvey's Legal Agent Benchmark - Environmental ESGLegal0.00.00.0
Harvey's Legal Agent Benchmark - Funds Asset ManagementLegal20.020.00.0
Harvey's Legal Agent Benchmark - Healthcare Life SciencesLegal20.020.00.0
Harvey's Legal Agent Benchmark - ImmigrationLegal0.00.00.0
Harvey's Legal Agent Benchmark - InsuranceLegal0.00.00.0
Harvey's Legal Agent Benchmark - International Trade SanctionsLegal0.00.00.0
Harvey's Legal Agent Benchmark - Litigation Dispute ResolutionLegal20.020.00.0
Harvey's Legal Agent Benchmark - Real EstateLegal0.00.00.0
Harvey's Legal Agent Benchmark - TaxLegal0.00.00.0
Harvey's Legal Agent Benchmark - Trusts Estates Private ClientLegal0.00.00.0
Harvey's Legal Agent Benchmark - White Collar Defense InvestigationsLegal0.00.00.0
OncologyBench Strict (25 tasks, October 9 rubric)Health & Medicine18.718.70.0

Grok 4.6vsGrok 4.7normalised score marginsData Nov 15, 2026Data: Vals AI, Artificial Analysis, Mercor and 57 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…