Atlas

Compare Models

Claude Opus 5 vs Grok 4.6

Claude Opus 5 leads by 4.1 capability points (≈95% interval +1.9 to +6.3): a clear lead over Grok 4.6 · P(higher) ≈ 100%

Wins 203 of 286 shared benchmarks (62 lost, 21 tied)

Capability vs. relative cost
Claude Opus 5Grok 4.6Closed 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.

Claude Opus 5vsGrok 4.6capability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins203 – 62 · 286 shared, 21 tied

Comparing matching setups only (24 benchmarks excluded) · suites counted once
← Claude Opus 5Grok 4.6 →

Claude Opus 5vsGrok 4.6wins by categoryData Nov 15, 2026Data: Vals AI, Mercor, benchmarks.bio and 83 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Grok 4.6 minus Claude Opus 5.

LLM Creative Story WritingChat & Writing3.7-2.9−74.7
Document ArenaProfessional Work15091438−63.1
NanoGPT Speedrun Frontier (human record gap closed)Agents53.610.1−58.4
ProofBench v1.1Math99.051.0−48.0
Medical Long Context Reasoning (MLCR-AA)Health & Medicine59.412.2−47.2
FrontierMath Tier 4 (v2) - PrivateMath73.231.7−41.5
MRCR v2 (8-needle, Context Arena) - 128K pointwiseGeneral QA91.352.6−38.7
IOI (Vals v2)Code84.347.6−36.7
BioSecBench-RefusalSafety29.464.2+34.8
LLM Chess EloGames1285443−34.6
CheatBenchSafety45.379.1−33.7
Antidote: Everyday EditionSafety1055966−31.8
Bio Capabilities Index (SecureBio)Indexes178137−30.0
Terminal-Bench 4.0Code53.925.3−28.6
ALE-BenchCode26221822−28.4
ARC-AGI-3 (Semi-Private)General QA30.22.1−28.0
LMArena Agent Arena Net ImprovementAgents9.51.1−27.7
KernelBench Hard Normalized (RTX PRO 6000)Code33.360.5+27.1
InferenceBenchCode8.96.7−26.8
FrontierSWE V2Code52.025.3−26.7
Wiz Cyber Model Arena v3.9Cybersecurity71.245.5−25.7
Mystery Game PuzzlesGames59.034.0−25.0
HealthBench ProfessionalHealth & Medicine73.448.5−24.9
Terminal-Bench 4.0 (Artificial Analysis)Code46.021.2−24.7
WeirdML (v2)Science91.867.3−24.5
ARC-AGI-2General QA90.467.1−23.3
Humanity's Last Exam DiamondGeneral QA44.821.5−23.3
MRCR v2 (8-needle, Context Arena) - 128K AUCGeneral QA97.574.4−23.1
BrowseComp Extended (Mercor held-out)Search44.021.5−22.5
LAB-Bench LitQA2 (SecureBio)Science51.428.9−22.5
Terminal-Bench-Science 0.1Science28.66.2−22.4
OncologyBench Strict (25 tasks, October 9 rubric)Health & Medicine40.018.7−21.3
Furniture Assembly BenchmarkMultimodal60.840.0−20.8
SWE Atlas Codebase QnA Extended (Mercor held-out)Code40.019.3−20.7
Olam Social Poker EloGames15181499−20.7
SAGEMath49.428.9−20.5
Harvey's Legal Agent Benchmark - Antitrust CompetitionLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Banking FinanceLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Corporate GovernanceLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Data Privacy CybersecurityLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Energy Natural ResourcesLegal20.040.0+20.0
Harvey's Legal Agent Benchmark - Funds Asset ManagementLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Litigation Dispute ResolutionLegal0.020.0+20.0
FrontierMath Tiers 1-3 (v2) - PrivateMath85.666.0−19.6
MedCodeHealth & Medicine63.644.7−18.9
WCB (SecureBio)Science68.850.3−18.5
HardGeoBench Avg DistanceMultimodal12181807−18.2
VoxelBench Text RatingGames23002050−17.8
Hemingway-benchChat & Writing1047983−17.7
Terminal-Bench 3 - release-reported accuracyCode43.326.0−17.3
LLM Debate BenchmarkChat & Writing17241604−17.1
Eval Connections One-Shot with TrapsGames74.091.0+17.0
HPCT (SecureBio)Science78.661.6−17.0
BuseyBench SVGMultimodal6.34.6−16.9
SWE-Marathon v1.1Code48.831.9−16.9
Conceptual Reasoning IndexIndexes74.762.5−16.7
HLE-Bio (SecureBio)Science45.929.2−16.7
DeepsecBench — Revalidated V1Code32.416.2−16.3
DeepsecBench — Revalidated V1 (233 issues)Code32.316.1−16.2
GameBench 2Games84.868.7−16.1
BlenderWorldCode73.057.0−16.0
LMCAGeneral QA64.548.5−16.0
Terminal-Bench Science 0.1 (Vals)Science27.111.4−15.7
MedXpertQA MM Extended (Mercor held-out)Health & Medicine64.148.4−15.7
NoCoT-Bench NCRI 15.2General QA11586.8−15.6
HLE Extended (Mercor held-out)General QA61.045.8−15.2
Co-Design BenchCode66.151.1−15.0
HackerRank-BenchCode42.627.7−14.9
Harvey LAB held-out set (Fireworks run)Legal7.422.1+14.7
SnorkelLegalLegal26.440.7+14.3
Extended NYT Connections 940 (Quadratic v1)Games94.380.0−14.3
LMArena Agent Arena Net Improvement (Steering Burden)Agents7.2-0.2−14.0
VariantBench V1 Clean RerunsScience52.839.0−13.8
BioSecBench-Refusal V1 Clean RerunsSafety31.845.6+13.8
HardGeoBench <=100miMultimodal46.332.5−13.7
LAB-Bench ProtocolQA (SecureBio)Science81.168.1−13.0
Time Horizon Index: KSPAgents18.85.8−13.0
Vending-Bench 2Games111829047−12.9
Code MigrationCode57.544.6−12.9
GBENCH One-shot Coding GScoreGames68.155.4−12.8
Artificial Analysis Coding Agent Index v1.5Indexes59.747.0−12.8
AA-AnalystAgent (pass^5)Professional Work53.841.3−12.5
HardGeoBench <=10miMultimodal32.520.0−12.5
Arena Image-to-WebDevCode16641607−12.4
ORCA-bench Hard RCA (Fireworks 324-task subset)Agents44.332.1−12.2
ACCoRDGeneral QA83.871.6−12.2
GlobeBenchGeneral QA15.928.1+12.2
Vibe Code Bench v1.1Code88.476.2−12.2
ScBench-Long V0 VerifiedScience36.424.2−12.1
YC-BenchGames19115921568241−12.0
MineBench — Bradley-Terry RatingGames20081811−11.9
Roboflow Vision Evals (October 2026, five tasks)Multimodal70.959.1−11.8
SCI-AGENT-EXAMScience64.252.4−11.8
Humanity's Last Exam (Text-Only)General QA54.442.9−11.5
FrontierCode 2.0 (Artificial Analysis)Code31.219.8−11.4
MetagenomicsBench V1Science53.342.0−11.3
Bedside BenchHealth & Medicine94.683.3−11.3
MetagenomicsBench V0Science53.342.0−11.3
Excel Modeling BenchmarkFinance & Accounting73.662.7−10.8
ScBench-Long V1 Clean RerunsScience27.316.5−10.8
Vals RSI Index v1.1 (historical five tasks)Agents32.121.5−10.6
AutomationBench-AA 1.0.6Agents56.667.0+10.4
Public Benefits Bench v1.1Professional Work76.966.8−10.1
ARC-AGI-1 Semi-PrivateGeneral QA97.587.5−10.0
Chess Puzzles (task version 2.0)Games43.033.0−10.0
HardGeoBench Country AccuracyMultimodal53.843.8−10.0
Vals Science IndexIndexes47.938.1−9.8
MBCT (SecureBio)Science60.250.4−9.8
RSI-Exam 0.1 (Full 88)Agents0.50.4−9.7
Roboflow Vision Evals (September 2026 protocol)Multimodal78.368.7−9.6
LMArena Code Arena WebDevCode16911618−9.6
CritPt (avg@5)Science29.119.7−9.4
CritPt v2 (agentic, avg@5)Science71.462.3−9.1
SUPERChemScience74.765.7−9.0
ZeroBench (Main questions, pass@5)Multimodal26.017.0−9.0
AA-LCR Extended (Mercor held-out)Search73.965.2−8.7
SnorkelFinance 2.0Finance & Accounting16.525.1+8.6
AutoResearchExam (hidden-test AUARC)Agents0.60.5−8.6
Senior SWE-BenchCode34.726.3−8.4
CyberArenaCybersecurity73.865.5−8.3
Snorkel SWE-bench CLICode14.05.7−8.3
MMMU-Pro Extended (Mercor held-out)Multimodal34.326.0−8.3
ReactBench (39 tasks)Code42.133.8−8.2
CadQueryEval (Corrected Geometry)Code80.072.0−8.0
Harvey LAB: In-House Contracting (held-out)Legal2.010.0+8.0
MortgageTax (Vals all 1,258 documents)Finance & Accounting72.164.2−7.9
KernelBench CUDA Normalized (RTX PRO 6000)Code45.537.7−7.8
Vals RSI Index v1.1 (four tasks)Agents33.025.5−7.6
Terminal-Bench 2.1 Extended (Mercor held-out)Code38.045.5+7.5
LMArena Agent Arena Praise vs ComplaintAgents12.2-2.5−7.3
Legal Research BenchLegal55.348.1−7.2
LiveBench 2.0Agents82.174.9−7.2
BioMysteryBench (Vals Terminus 2)Science79.372.2−7.0
CorpFin v2Finance & Accounting73.266.2−7.0
GeneBench-Pro Public Release SubsetScience22.015.0−7.0
WordleBenchGames96.089.0−7.0
SnorkelManufacturingProfessional Work9.416.4+7.0
GDP.pdfProfessional Work24.017.2−6.8
MysteryMechanismScience37.430.6−6.8
DeepSWE v1.1Code73.266.7−6.4
LMArena Agent Arena Confirmed SuccessAgents10.2-2.6−6.4
SpatialBench V0 VerifiedScience71.077.4+6.4
ZeroBench Main Mean pass@1 (five samples)Multimodal17.211.0−6.2
BioSecBench-Function V0Science50.344.1−6.2
ATLAS-Finance v1.1Finance & Accounting10.74.7−6.0
APEX-SWECode63.757.7−6.0
Toloka Arena (v3)Agents62.056.0−6.0
𝜏³-BankingAgents44.750.7+6.0
LMArena Text Arena EloChat & Writing14901456−5.8
LAB-Bench CloningScenarios (SecureBio)Science100.094.2−5.8
BioSecBench-Structure V1 (58 evaluations)Science20.714.9−5.7
Vision Arena OverallMultimodal12851263−5.7
HieroglyphBenchMultimodal14.620.3+5.7
DAYJOB: HealthcareHealth & Medicine8.42.8−5.6
Design Arena Overall Code CategoriesCode13211288−5.5
ChessBench CoherenceGames63.869.2+5.4
Epoch Capabilities Index (ECI)Indexes164158−5.3
TxBench Oligo Discovery V0Science48.142.8−5.3
Snorkel Agentic Coding 2.0Code38.933.6−5.3
NoCoT-Bench NCKI (kspine v3)General QA119112−5.3
Tax Agent BenchFinance & Accounting76.771.4−5.3
CursorBench 4.0Code46.641.4−5.2
MacroscopeBenchCode69.975.1+5.2
Arena Alignment IndexSafety75.680.8+5.2
HANDBOOK.mdProfessional Work32.327.3−5.0
Harvey HLAB — Criteria Pass RateLegal87.792.5+4.8
GPQA-Bio (SecureBio)Science82.787.4+4.7
SimpleBenchGeneral QA80.675.9−4.7
AA-Briefcase EloAgents16251545−4.6
Biopharma Bench V0.1Health & Medicine67.663.0−4.6
MedXpertQA MM (200-question set)Health & Medicine79.875.2−4.6
RedlineBenchLegal44.048.5+4.5
MedScribeHealth & Medicine91.086.5−4.5
SciUniverse Level 1Science30.526.2−4.3
Artificial Analysis Progress Index v1Indexes334322−4.3
CVDP (517-problem set)Code62.258.0−4.2
TaxEval v2Finance & Accounting75.171.1−4.0
PG-LLM-50Science0.40.3−4.0
APEX-Accounting 1.1Finance & Accounting64.360.3−4.0
BabaIsBench Stage 0: The IntroGames100.096.0−4.0
BioSecBench-Surveillance V0 (102 tasks)Science51.547.5−4.0
MBABench 2.0Finance & Accounting4.00.0−4.0
SWE-bench Verified Extended (Mercor held-out)Code82.078.0−4.0
ZeroBench Main Questions pass^5Multimodal11.07.0−4.0
BioSecBench-Surveillance V1 Clean RerunsScience51.547.5−4.0
LMArena Agent Arena Bash RecoveryAgents13.65.7−3.9
VIALSMultimodal24.420.5−3.9
TxBench Oligo Discovery V1 Clean RerunsScience45.641.7−3.9
ChessBench EloGames826891+3.9
VariantBenchScience49.746.0−3.7
LMArena Agent Arena SteerabilityAgents10.73.3−3.7
SnorkelRevOpsProfessional Work12.215.8+3.6
SpatialBench-Long V1 Clean RerunsScience30.927.4−3.6
MMMU ProMultimodal84.080.5−3.5
ChessBench AccuracyGames71.067.8−3.4
PWNBench v0.1Cybersecurity60.063.3+3.4
AA-Omniscience IndexIndexes37.130.5−3.3
SM BenchSafety85.889.0+3.2
SWE-TogetherCode63.860.6−3.2
Toloka Arena (v1)Agents52.949.7−3.2
Concept Synth INDUCTION Challenge64 Round 1General QA32.835.9+3.1
EnterpriseBench: CoreCraftProfessional Work68.765.6−3.1
VCT (SecureBio)Science51.448.3−3.1
FrontierChallenge (Pass Rate)Science17.520.6+3.1
LMArena Agent Arena Steering BurdenAgents4.4-1.8−3.1
GPQA DiamondGeneral QA92.495.5+3.0
BioSecBench-RespiratoryVirus-Surveillance V1 (110 evaluations)Science14.611.5−3.0
SpeciEval (Pooled English Runs)Safety87.990.8+2.9
SpeciEval (Refusal-Aware Retries, Historical)Safety87.990.8+2.9
Blueprint-Bench 2Multimodal0.30.3+2.8
micro1 PharmaBenchHealth & Medicine9.712.5+2.8
Realm Financial ReasoningFinance & Accounting47.444.6−2.8
SciCode (main problems)Science17.314.6−2.7
TxBench Antibody Discovery V1 Clean RerunsScience41.944.5+2.7
ProgramBench (Mercor 100-task subset)Code4.31.7−2.6
MMLU ProGeneral QA92.089.4−2.6
FrontierCode 1.1 Extended - Pass RateCode69.667.0−2.6
Hedge-BenchFinance & Accounting29.026.4−2.5
SWEeper-BenchCode56.559.0+2.5
CharXiv Extended (Mercor held-out)Multimodal79.877.3−2.5
GDPval-AA v2Professional Work17081663−2.2
Snorkel WorkplaceAgentsProfessional Work14.917.1+2.2
RuneBenchGames5.75.8+2.0
TxBench Antibody Discovery V0Science53.051.0−2.0
CharXiv (descriptive and reasoning combined)Multimodal94.692.9−1.7
SAB ResearchFinance & Accounting96.294.6−1.6
GDPval-AA v2.1Professional Work16951658−1.6
Bug Hunt BenchCode25.727.3+1.6
EpiBench V1 Clean RerunsScience25.223.6−1.6
BioSecBench-Function V1 Clean RerunsScience42.544.0+1.6
SpatialBench-Long V0 VerifiedScience28.830.3+1.5
DAYJOB: FinanceFinance & Accounting11.312.8+1.5
SWE-bench VerifiedCode97.095.6−1.4
TxBench Preclinical Pharmacology V1 Clean RerunsScience63.061.7−1.3
Snorkel Digital Work IndexIndexes21.820.5−1.2
SpatialBench V1 Clean RerunsScience68.467.3−1.2
UnityBench-Mini3DCode71.870.6−1.1
GDPval Extended (Mercor held-out rubric score)Professional Work62.961.8−1.1
AA-LCR v1.1Search82.081.0−1.0
Concept Synth INDUCTION Challenge100 Round 1General QA24.025.0+1.0
CrossSearch (pass^3)Search34.035.0+1.0
ForecastBench Preliminary Dataset Brier IndexForecasting62.061.0−1.0
EpiBenchScience35.234.3−0.9
APEX-1 (Mercor, mean score)Professional Work68.667.7−0.9
EnterpriseOps-Gym-AAAgents47.548.3+0.9
ScBench V1 Clean RerunsScience58.359.1+0.9
LiveCodeBench (Vals AI public dataset)Code89.088.2−0.8
ComplexConstraintsChat & Writing37.336.5−0.8
ForecastBench Tournament Overall Brier IndexForecasting66.665.8−0.8
Clipping Bench v1.0Games734805+0.8
Terminal-Bench 2.1Code87.688.4+0.7
MMLU-Bio (SecureBio)Science91.191.8+0.7
LegalBenchLegal87.086.3−0.7
BrowseComp (Mercor 130-question subset)Search84.684.0−0.6
DTBenchGeneral QA97.997.3−0.5
SnorkelUnderwrite 2.0Finance & Accounting18.317.8−0.5
AdvancedIFChat & Writing80.580.9+0.4
GDPval (Mercor 213-task rubric score)Professional Work85.686.0+0.4
Artificial Analysis Intelligence IndexIndexes60.761.0+0.3
APEX-Agents 1.1Professional Work77.077.3+0.3
OTIS Mock AIME 2024-2025Math98.999.2+0.3
LMArena Agent Arena Bash ReliabilityAgents9.810.0+0.1
SciCode 1.0.1 (Artificial Analysis)Code56.456.5+0.1
CursorBench 3.2Code70.069.9−0.1
LMArena Agent Arena Tool HallucinationAgents0.91.00.0
AIME 2025Math100.0100.00.0
DAXBench v2Code86.286.20.0
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 - Capital MarketsLegal0.00.00.0
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 - 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 - Intellectual PropertyLegal0.00.00.0
Harvey's Legal Agent Benchmark - International Trade SanctionsLegal0.00.00.0
Harvey's Legal Agent Benchmark - Real EstateLegal0.00.00.0
Harvey's Legal Agent Benchmark - Structured Finance SecuritizationLegal20.020.00.0
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
ObviousBenchGeneral QA100.0100.00.0
Shroombench (mushroom identification, top-1)Multimodal44.044.00.0

Claude Opus 5vsGrok 4.6normalised score marginsData Nov 15, 2026Data: Vals AI, Mercor, benchmarks.bio and 83 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…