Atlas

Compare Models

Claude Opus 5.5 vs Grok 4.6

Claude Opus 5.5 leads by 7.7 capability points (≈95% interval +5.4 to +10.1): a clear lead over Grok 4.6 · P(higher) ≈ 100%

Wins 188 of 244 shared benchmarks (39 lost, 17 tied)

Capability vs. relative cost
Claude Opus 5.5Grok 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 5.5vsGrok 4.6capability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins188 – 39 · 244 shared, 17 tied

Comparing matching setups only (19 benchmarks excluded) · suites counted once
← Claude Opus 5.5Grok 4.6 →

Claude Opus 5.5vsGrok 4.6wins by categoryData Nov 15, 2026Data: Vals AI, Mercor, Artificial Analysis and 73 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 5.5.

KernelBench Mega Normalized (RTX PRO 6000)Code100.00.0−100.0
Time Horizon Index: KSPAgents91.35.8−85.5
AA-PersonalityBench Imagination indexChat & Writing94.320.0−75.0
LLM Creative Story WritingChat & Writing3.7-2.9−74.6
CheatBenchSafety7.279.6−72.4
AA-PersonalityBench Organization indexChat & Writing79.114.3−65.4
FrontierMath Tier 4 (v2) - PrivateMath95.031.7−63.3
Terminal-Bench-Science 0.1Science61.92.9−59.0
Medical Long Context Reasoning (MLCR-AA)Health & Medicine66.712.2−54.4
AA-PersonalityBench Warmth indexChat & Writing84.831.5−53.8
Gray Swan IPI Attack Success (k=15)Safety1.051.8−50.8
ProofBench v1.1Math100.051.0−49.0
LLM Chess EloGames1575443−46.5
AA-PersonalityBench Informativeness indexChat & Writing90.544.8−46.2
YC-BenchGames28737251568241−45.6
Terminal-Bench Science 0.1 (Vals)Science48.64.3−44.3
Furniture Assembly BenchmarkMultimodal83.340.0−43.3
BioSecBench-Refusal V1 Clean RerunsSafety3.846.9+43.1
LAB-Bench LitQA2 (SecureBio)Science71.128.9−42.2
TxBench Oligo Discovery V1 Clean RerunsScience1.443.3+41.9
VoxelBench Text RatingGames26132050−40.1
Concept Synth INDUCTION Challenge100 Round 1General QA65.025.0−40.0
CAD ArenaAgents74.634.8−39.8
Terminal-Bench 4.0Code64.825.3−39.5
FrontierSWE V2 Worst@5Code52.012.9−39.1
BrowseComp Extended (Mercor held-out)Search60.521.5−39.0
KernelBench CUDA Normalized (RTX PRO 6000)Code76.337.7−38.5
MirrorCode (ML, +Private, 2L)Code78.941.1−37.8
TxBench Antibody Discovery V1 Clean RerunsScience6.844.5+37.7
LMArena Agent Arena Net Improvement (Steering Burden)Agents19.5-0.2−37.2
Mystery Game PuzzlesGames71.034.0−37.0
Concept Synth INDUCTION Challenge64 Round 1General QA71.935.9−35.9
Terminal-Bench 4.0 (Artificial Analysis)Code56.621.2−35.4
TxBench Preclinical Pharmacology V1 Clean RerunsScience26.361.7+35.3
HardGeoBench Avg DistanceMultimodal6691807−35.1
FrontierSWE V2 Best@5Code71.537.3−34.2
Chess Puzzles (task version 2.0)Games67.033.0−34.0
Humanity's Last Exam DiamondGeneral QA54.621.5−33.1
ALE-BenchCode27481822−32.8
TCSExam LiteMath72.342.1−30.2
BioSecBench-Function V1 Clean RerunsScience73.644.0−29.6
HardGeoBench <=100miMultimodal61.332.5−28.8
CadQueryEval (Corrected Geometry)Code100.072.0−28.0
ChessBench EloGames1349891−27.2
AA-PersonalityBench Expressiveness indexChat & Writing90.563.8−26.9
ARC-AGI-2General QA93.367.1−26.2
GameBench 2Games94.868.7−26.1
ZeroBench (Main questions, pass@5)Multimodal43.017.0−26.0
LAB-Bench ProtocolQA (SecureBio)Science94.068.1−25.9
LMArena Code Arena WebDevCode18141618−25.7
FrontierMath Tiers 1-3 (v2) - PrivateMath91.266.0−25.3
MMMU-Pro Extended (Mercor held-out)Multimodal51.026.0−25.0
Bio Capabilities Index (SecureBio)Indexes171137−24.5
Design Creativity BenchCode51.043.4−24.2
ZeroBench Main Mean pass@1 (five samples)Multimodal35.011.0−24.0
SWE Atlas Codebase QnA Extended (Mercor held-out)Code43.319.3−24.0
FrontierCode 2.0 (Artificial Analysis)Code43.819.8−24.0
HardGeoBench Country AccuracyMultimodal67.543.8−23.7
HackerRank-BenchCode51.227.7−23.6
AA-PersonalityBench Conviction indexChat & Writing69.592.4+23.1
HealthBench ProfessionalHealth & Medicine71.448.5−22.9
Roboflow Vision Evals (October 2026, five tasks)Multimodal82.059.1−22.9
SimpleQA Verified (Epoch anti-abstention)General QA72.249.3−22.9
MedXpertQA MM Extended (Mercor held-out)Health & Medicine71.248.4−22.8
HardGeoBench <=10miMultimodal42.520.0−22.5
BuseyBench SVGMultimodal6.84.6−22.4
ChessBenchGames65.343.3−22.0
CrossSearch (pass^3)Search57.035.0−22.0
DAYJOB: HealthcareHealth & Medicine24.72.8−21.9
LMArena Agent Arena Praise vs ComplaintAgents40.3-2.5−21.4
Conceptual Reasoning IndexIndexes77.962.5−21.1
ReactBench (39 tasks)Code54.433.8−20.5
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 - Corporate M&ALegal0.020.0+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
Harvey's Legal Agent Benchmark - Structured Finance SecuritizationLegal0.020.0+20.0
OncologyBench Strict (25 tasks, October 9 rubric)Health & Medicine37.317.3−20.0
LMCAGeneral QA68.248.5−19.8
VIALSMultimodal40.220.5−19.7
MineBench — Bradley-Terry RatingGames21341811−19.6
AA-PersonalityBench Transparency indexChat & Writing90.571.4−19.2
CritPt v2 (agentic, avg@5)Science81.462.3−19.1
Artificial Analysis Coding Agent Index v1.5Indexes66.047.0−19.0
Toloka Arena (v3)Agents75.056.0−19.0
GBENCH One-shot Coding GScoreGames73.955.4−18.5
MBCT (SecureBio)Science68.750.4−18.3
ProgramBench (Mercor 100-task subset)Code20.01.7−18.3
Blueprint-Bench 2Multimodal0.50.3−18.0
ZeroBench Main Questions pass^5Multimodal25.07.0−18.0
Hedge-BenchFinance & Accounting44.226.4−17.7
HieroglyphBenchMultimodal37.420.3−17.1
SAGEMath45.828.9−16.9
RSI-Exam 0.1 (Full 88)Agents0.50.4−16.9
ACCoRDGeneral QA88.471.6−16.8
Roboflow Vision Evals (September 2026 protocol)Multimodal85.568.7−16.8
Harvey HLAB — Criteria Pass RateLegal75.892.5+16.7
CursorBench 4.0Code57.841.4−16.4
BioSecBench-Structure V1 (58 evaluations)Science31.014.9−16.1
Design Arena Overall Code CategoriesCode13841288−15.9
Harvey LAB held-out set (Fireworks run)Legal6.322.1+15.8
Vals RSI Index v1.1 (historical five tasks)Agents37.121.5−15.6
EpiBench V1 Clean RerunsScience39.023.6−15.4
AA-PersonalityBench Slop indexChat & Writing20.035.3+15.4
Veracode GenAI Code Security (Fall 2026)Cybersecurity73.058.0−15.0
AA-AnalystAgent (pass^5)Professional Work56.341.3−15.0
GeneBench-Pro Public Release SubsetScience30.015.0−15.0
Humanity's Last Exam (Text-Only)General QA57.542.9−14.6
SWE-Marathon v1.1Code46.331.9−14.4
GDPval Extended (Mercor held-out rubric score)Professional Work75.861.8−14.0
BioSecBench-RespiratoryVirus-Surveillance V1 (110 evaluations)Science25.411.5−13.9
SUPERChemScience79.665.7−13.9
LLM Debate BenchmarkChat & Writing17011604−13.8
SWE-bench VerifiedCode98.284.6−13.6
ComplexConstraintsChat & Writing49.936.5−13.4
Excel Modeling BenchmarkFinance & Accounting75.962.7−13.2
VCT (SecureBio)Science61.448.3−13.1
Clipping Bench v1.0Games2018805−13.0
Senior SWE-BenchCode39.326.3−13.0
RuneBenchGames6.75.8−12.7
SciCode (main problems)Science27.314.6−12.7
SimpleBenchGeneral QA88.475.9−12.5
Bug Hunt BenchCode39.727.3−12.4
ChessBench AccuracyGames79.367.8−12.3
AA-Briefcase v1.1 EloAgents17611540−12.1
CritPt (avg@5)Science31.719.7−12.0
SWE-bench Verified Extended (Mercor held-out)Code66.078.0+12.0
Vals RSI Index v1.1 (four tasks)Agents37.325.5−11.9
LMArena Agent Arena Confirmed SuccessAgents20.7-2.6−11.6
AutoResearchExam (hidden-test AUARC)Agents0.60.5−11.6
LiveBench 2.0Agents86.574.9−11.6
DAYJOB: FinanceFinance & Accounting23.912.8−11.1
GDP.pdf (Artificial Analysis, All-pass)Professional Work28.817.8−11.0
ARC-AGI-1 Semi-PrivateGeneral QA98.587.5−11.0
Snorkel SWE-bench CLICode16.65.7−10.9
ChessBench CoherenceGames80.169.2−10.9
DeepsecBench — Revalidated V1Code26.716.2−10.5
MedXpertQA MM (200-question set)Health & Medicine85.775.2−10.5
BioSecBench-Surveillance V1 Clean RerunsScience58.548.0−10.5
DeepsecBench — Revalidated V1 (233 issues)Code26.616.1−10.5
SciCode 1.0.1 (Artificial Analysis)Code66.956.5−10.4
LMArena Agent Arena Steering BurdenAgents18.6-1.8−10.2
VariantBench V1 Clean RerunsScience50.640.4−10.2
CWE-bench v1 Programmatic Pass@1Cybersecurity67.057.0−10.0
WellnessBenchSafety53.043.0−10.0
APEX-SWECode67.657.7−9.9
DAXBench v2Code95.786.2−9.6
SnorkelLegalLegal31.440.7+9.3
τ³-Banking (Mercor run, bm25_grep retrieval)Agents55.045.7−9.3
LibraryDesignBenchCode48.939.7−9.2
AA-LCR Extended (Mercor held-out)Search74.265.2−9.0
EnterpriseBench: CoreCraftProfessional Work74.465.6−8.8
Epoch Capabilities Index (ECI)Indexes169158−8.7
LMArena Text Arena EloChat & Writing15061456−8.7
Extended NYT Connections 940 (Quadratic v1)Games88.580.0−8.5
SWE-TogetherCode68.860.6−8.2
CloudOpsBench v1.0Agents47.839.7−8.1
AA-Omniscience IndexIndexes46.430.5−8.0
GDPval-AA v2.1Professional Work18361658−7.8
SpatialBench-Long V1 Clean RerunsScience19.627.4+7.7
Olam Social Poker EloGames15061499−7.6
CharXiv Extended (Mercor held-out)Multimodal84.777.3−7.4
SM BenchSafety81.889.0+7.2
Harvey LAB: In-House Contracting (held-out)Legal3.010.0+7.0
Snorkel Agentic Coding 2.0Code40.433.6−6.8
DeepSWE v1.1Code74.267.5−6.7
RedlineBenchLegal42.048.5+6.5
SnorkelFinance 2.0Finance & Accounting18.625.1+6.5
Immigration BenchLegal72.166.0−6.2
CVDP (517-problem set)Code64.158.0−6.1
MMMU ProMultimodal86.680.5−6.1
xbench-ScienceQAScience76.470.4−6.0
SnorkelManufacturingProfessional Work10.416.4+6.0
AHK-EvalCode91.797.2+5.6
SpatialBench V1 Clean RerunsScience72.867.3−5.5
τ³-Banking (Fireworks, banking_knowledge)Agents56.050.7−5.3
APEX-Accounting 1.1Finance & Accounting65.560.3−5.2
MedCodeHealth & Medicine49.844.7−5.1
Eval Connections One-Shot with TrapsGames86.091.0+5.0
Finance Agent v2Finance & Accounting58.653.7−4.9
SnorkelUnderwrite 2.0Finance & Accounting22.717.8−4.9
MedScribeHealth & Medicine91.486.5−4.9
Terminal-Bench 2.1Code83.588.4+4.9
LMArena Agent Arena Bash RecoveryAgents15.05.7−4.7
Biopharma Bench V0.1Health & Medicine67.563.0−4.5
BrowseComp (Mercor 130-question subset)Search88.584.0−4.5
WCB (SecureBio)Science54.550.3−4.2
APEX-Agents 1.1Professional Work81.377.3−4.0
MMLU ProGeneral QA93.489.4−4.0
FrontierCode 1.1 Extended - ScoreCode65.361.3−4.0
ScBench V1 Clean RerunsScience55.459.1+3.8
FrontierCode 1.1 Extended - Pass RateCode70.867.0−3.7
AA-LCR v1.1Search84.781.0−3.7
SpeciEval (Pooled English Runs)Safety87.290.8+3.6
SpeciEval (Refusal-Aware Retries, Historical)Safety87.290.8+3.6
CharXiv (descriptive and reasoning combined)Multimodal96.492.9−3.5
GPQA-Bio (SecureBio)Science90.987.4−3.5
SnorkelRevOpsProfessional Work12.515.8+3.3
Snorkel Digital Work IndexIndexes23.620.5−3.1
MacroscopeBenchCode78.975.9−3.0
SWE-bench MultilingualCode95.592.7−2.8
Snorkel WorkplaceAgentsProfessional Work14.417.1+2.7
SAB ResearchFinance & Accounting97.294.6−2.7
LMArena Agent Arena Bash ReliabilityAgents15.110.0−2.5
Access to Justice BenchmarkLegal73.871.3−2.4
Arena Alignment IndexSafety83.280.8−2.4
GDPval (Mercor 213-task rubric score)Professional Work88.386.0−2.3
ScBench-Long V1 Clean RerunsScience13.615.9+2.3
Terminal-Bench 2.1 Extended (Mercor held-out)Code47.545.5−2.0
APEX-1 (Mercor, mean score)Professional Work65.867.7+1.9
AdvancedIFChat & Writing82.880.9−1.9
GlobeBenchGeneral QA26.328.1+1.8
MetagenomicsBench V0Science43.742.0−1.7
micro1 PharmaBenchHealth & Medicine10.812.5+1.7
MetagenomicsBench V1Science43.742.0−1.7
DTBenchGeneral QA98.997.3−1.6
SWE Atlas - Codebase QnACode48.950.3+1.4
ForecastBench Preliminary Dataset Brier IndexForecasting62.261.0−1.2
MMLU-Bio (SecureBio)Science92.791.8−0.9
OTIS Mock AIME 2024-2025Math100.099.2−0.8
Prophet Arena Arena Score (Agentic, OpenCode)Forecasting86.285.6−0.6
Prophet Arena Arena Score (Fixed Context)Forecasting85.585.0−0.5
LMArena Agent Arena Tool HallucinationAgents0.71.0+0.2
AIME 2025Math100.0100.00.0
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 - Capital MarketsLegal0.00.00.0
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 - 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

Claude Opus 5.5vsGrok 4.6normalised score marginsData Nov 15, 2026Data: Vals AI, Mercor, Artificial Analysis and 73 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…