Atlas

Compare Models

Claude Opus 5.5 vs Grok 4.7

Claude Opus 5.5 leads by 7.1 capability points (≈95% interval +4.7 to +9.6): a clear lead over Grok 4.7 · P(higher) ≈ 100%

Wins 244 of 313 shared benchmarks (48 lost, 21 tied)

Capability vs. relative cost
Claude Opus 5.5Grok 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.

Claude Opus 5.5vsGrok 4.7capability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins244 – 48 · 313 shared, 21 tied

Comparing matching setups only (8 benchmarks excluded) · suites counted once
← Claude Opus 5.5Grok 4.7 →

Claude Opus 5.5vsGrok 4.7wins by categoryData Nov 15, 2026Data: Vals AI, Mercor, Artificial Analysis and 94 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 Claude Opus 5.5.

KernelBench Mega Normalized (RTX PRO 6000)Code100.018.0−82.0
FrontierMath Tier 4 (v2) - PrivateMath95.017.1−77.9
ProofBench v1.1Math100.026.0−74.0
CheatBenchSafety7.278.0−70.8
Furniture Assembly BenchmarkMultimodal83.320.8−62.5
BioSecBench-Refusal V1 Clean RerunsSafety0.062.4+62.4
AA-PersonalityBench Imagination indexChat & Writing94.335.3−59.6
AA-PersonalityBench Informativeness indexChat & Writing90.533.4−57.7
BrokenArXiv 05/2026Math95.040.0−55.0
WARC-Bench (192-task test split)Agents90.335.6−54.7
LAB-Bench LitQA2 (SecureBio)Science71.116.9−54.2
HoneyBench v0.2 (hack rate)Safety31.685.1−53.5
Medical Long Context Reasoning (MLCR-AA)Health & Medicine66.715.0−51.7
Terminal-Bench-Science 0.1Science61.911.4−50.5
AA-PersonalityBench Organization indexChat & Writing79.129.6−50.0
BrokenArXiv 09/2026 (Native Harnesses)Math76.727.7−49.1
CAD ArenaAgents74.625.8−48.8
FrontierML ($200 wallet)Agents58.810.9−47.9
BrokenArXiv 06/2026Math100.052.8−47.2
ViBench 2.0Code48.51.5−47.0
TreasuryQA BenchProfessional Work64.818.1−46.7
HardGeoBench Avg DistanceMultimodal6692104−44.3
Toloka Arena (v3)Agents75.031.0−44.0
LAB-Bench ProtocolQA (SecureBio)Science94.051.4−42.6
BrokenArXiv 08/2026 (Native Harnesses)Math77.435.1−42.3
YC-BenchGames28737251666389−42.1
Mystery Game PuzzlesGames71.029.0−42.0
Concept Synth INDUCTION Challenge100 Round 1General QA65.024.0−41.0
AA-PersonalityBench Warmth indexChat & Writing84.844.8−40.4
Harvey's Legal Agent Benchmark - Capital MarketsLegal0.040.0+40.0
ArXivMath 08/2026 (Native Harnesses)Math86.046.5−39.5
TxBench Oligo Discovery V1 Clean RerunsScience1.140.3+39.2
FrontierMath Tiers 1-3 (v2) - PrivateMath91.253.0−38.2
Concept Synth INDUCTION Challenge64 Round 1General QA71.934.4−37.5
Terminal-Bench Science 0.1 (Vals)Science48.611.4−37.1
LLM Creative Story WritingChat & Writing3.70.4−37.1
ArXivMath 09/2026 (Native Harnesses)Math82.145.3−36.8
TxBench Antibody Discovery V1 Clean RerunsScience6.142.8+36.7
FrontierSWE V2 Worst@5Code52.015.9−36.1
Terminal-Bench 2.1 Extended (Mercor held-out)Code47.511.8−35.7
Design Creativity BenchCode51.062.2+35.7
FrontierMemory v1.1Agents73.337.8−35.6
OpenWorldEngineer (Mean@8)Code41.67.0−34.6
TaxCalcBench TY25Finance & Accounting42.08.0−34.0
MatFlowBench Pass@1Science56.722.8−33.9
Terminal-Bench 4.0 (Artificial Analysis)Code59.625.8−33.8
HardGeoBench <=100miMultimodal61.327.5−33.7
Agent Initiative BenchmarkAgents7.84.5−33.3
Code2Video-AA EloCode15361127−33.2
ALE-BenchCode27481823−32.8
AA-PersonalityBench Transparency indexChat & Writing90.558.1−32.7
LMArena Agent Arena Net Improvement (Steering Burden)Agents19.52.6−32.0
ARC-AGI-2General QA93.361.4−31.9
WitnessBench v2.0Games46.514.8−31.7
Web Clone BenchCode34.22.8−31.4
KernelBench CUDA Normalized (RTX PRO 6000)Code76.345.2−31.1
Humanity's Last Exam DiamondGeneral QA54.623.6−31.0
LLM Chess EloGames1575822−30.9
DAYJOB: Data ScienceProfessional Work37.36.7−30.6
AA-3D v1.0 EloCode1202992−30.5
FrontierSWE V2 Best@5Code71.541.1−30.5
CritPt v2 (agentic, avg@5)Science81.451.4−30.0
RWE-benchAgents29.00.0−29.0
Real-SWE 1.1Code56.828.0−28.8
SWE Atlas Codebase QnA Extended (Mercor held-out)Code43.315.3−28.0
SWE-bench Multimodal (Vals v1, 480 Tasks)Code70.242.9−27.3
Terminal-Bench 4.0Code64.837.6−27.3
sudo L7Code41.715.6−26.1
VulcanBench Frontier v4 (functional pass@1)Code73.9100.0+26.1
Chess Puzzles (task version 2.0)Games67.041.0−26.0
HardGeoBench <=10miMultimodal42.517.5−25.0
TxBench Preclinical Pharmacology V1 Clean RerunsScience30.754.7+24.0
HardGeoBench Country AccuracyMultimodal67.543.8−23.7
o11y-bench 1.1 Pass^3Agents93.870.3−23.4
FeatureArc (whole-feature chain gate)Code49.326.0−23.3
FrontierPhysics v0.1Science27.44.2−23.2
DistressBenchSafety75.652.4−23.2
BuseyBench SVGMultimodal6.84.6−22.8
LMArena Code Arena WebDevCode18141641−22.7
MBCT (SecureBio)Science68.746.1−22.6
SAFE-Teen 106 worst branch: Relationships (no prompt)Safety25.347.9−22.6
GameBench 2Games94.872.3−22.4
τ³-Banking (Mercor run, bm25_grep retrieval)Agents55.032.6−22.4
RiemannBench (November 2026 rescoring)Math80.057.6−22.4
ReactBench (39 tasks)Code54.432.3−22.1
VIALSMultimodal40.218.2−22.0
Bio Capabilities Index (SecureBio)Indexes171141−21.5
BrowseComp Extended (Mercor held-out)Search60.539.5−21.0
ArXivMath 05/2026Math81.760.8−20.8
DeepFinanceFinance & Accounting54.834.1−20.7
Conceptual Reasoning IndexIndexes77.963.1−20.3
GeneBench-Pro Public Release SubsetScience30.010.0−20.0
Harvey's Legal Agent Benchmark - Antitrust CompetitionLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Corporate M&ALegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Funds Asset ManagementLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Intellectual PropertyLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Litigation Dispute ResolutionLegal0.020.0+20.0
ProgramBench (Mercor 100-task subset)Code20.00.0−20.0
FrontierCode 2.0 - Pass RateCode51.832.3−19.5
SAFE-Teen Relationships (no prompt)Safety13.032.5−19.5
SAFE-Teen 106: Relationships (no prompt)Safety12.231.5−19.3
RuneBenchGames6.75.4−19.2
MedXpertQA MM Extended (Mercor held-out)Health & Medicine71.252.0−19.2
SWE-bench VerifiedCode98.279.1−19.1
Design Arena Overall Code CategoriesCode13841270−18.9
LMCAGeneral QA68.249.4−18.8
Short Form Bench v1.0Games23892578+18.7
Blueprint-Bench 2Multimodal0.50.3−18.7
OncologyBench Strict (25 tasks, October 9 rubric)Health & Medicine37.318.7−18.7
LMArena Agent Arena Praise vs ComplaintAgents40.33.1−18.6
PLE (Physical-Engineering's Last Exam)Professional Work50.632.0−18.6
Roboflow Vision Evals (October 2026, five tasks)Multimodal82.063.5−18.5
VCT v2 (SecureBio)Science60.542.0−18.5
FrontierCode 2.0 - ScoreCode47.328.9−18.4
Physical-Engineering's Last Exam (hidden test)Professional Work48.630.2−18.4
BioSecBench-Function V1 Clean RerunsScience61.543.3−18.2
SAFE-Teen 106 worst branch: Urgent Safety (no prompt)Safety26.544.4−17.9
EpiBench V1 Clean RerunsScience38.720.8−17.9
GBENCH One-shot Coding GScoreGames73.956.3−17.6
Clipping Bench v1.0Games2018384−17.6
MineBench — Bradley-Terry RatingGames21341848−17.3
Vals RSI Index v1.1 (four tasks)Agents37.320.2−17.1
MedXpertQA MM (200-question set)Health & Medicine85.769.0−16.7
SAFE-Teen Health Advice (no prompt)Safety32.816.1+16.7
DAYJOB: HealthcareHealth & Medicine24.78.4−16.3
SAFE-Teen 106: Health Advice (no prompt)Safety35.018.7+16.3
ChessBench EloGames13491076−16.2
SimpleQA Verified (Epoch anti-abstention)General QA72.256.0−16.2
SAFE-Teen 106 worst branch: Urgent Safety (under-18 prompt)Safety10.526.7−16.1
FrontierCUA pass@1Agents36.020.0−16.0
DeepsecBench — Revalidated V1Code26.710.9−15.7
DeepsecBench — Revalidated V1 (233 issues)Code26.610.9−15.7
AA-PersonalityBench Expressiveness indexChat & Writing90.575.3−15.4
SWE Atlas v1.1 - OverallCode77.462.1−15.3
RSI-Exam 0.1 (Full 88)Agents0.50.4−15.1
APEX-AccountingFinance & Accounting61.846.7−15.1
Harvey HLAB — Criteria Pass RateLegal75.890.5+14.7
DAYJOB: AccountingFinance & Accounting34.420.0−14.4
Robo-benchAgents14.60.2−14.4
ACCoRDGeneral QA88.474.0−14.4
Humanity's Last Exam (Text-Only)General QA57.543.1−14.4
APEX-SWECode67.653.4−14.2
BioResearchBenchScience18.04.0−14.0
CritPt (avg@5)Science31.718.0−13.7
Roboflow Vision Evals (September 2026 protocol)Multimodal85.571.9−13.6
AA-PersonalityBench Conviction indexChat & Writing69.582.9+13.5
ArXivMath 06/2026Math83.370.1−13.2
BrowseComp (Mercor 130-question subset)Search88.575.8−12.7
APEX-Agents 1.1Professional Work81.368.7−12.6
LiveBench 2.0Agents86.574.1−12.4
CWE-bench v2 Pass@1Cybersecurity67.154.7−12.4
AgenticRL-Bench (Verifiable)Agents14.01.8−12.3
Westworld Finance Diligence BenchFinance & Accounting59.447.2−12.2
Bug Hunt BenchCode39.727.6−12.1
CadQueryEval (Corrected Geometry)Code100.088.0−12.0
Senior SWE-BenchCode39.327.4−11.9
Extended NYT Connections 940 (Quadratic v1)Games88.576.8−11.7
LLM Debate BenchmarkChat & Writing17011620−11.5
AA-PersonalityBench Slop indexChat & Writing20.08.6−11.5
CursorBench 4.0Code57.846.3−11.5
SWE-bench MultilingualCode95.584.2−11.3
LMArena Text Arena EloChat & Writing15061442−11.2
MentalHealthBenchSafety52.441.3−11.1
SWE-bench Verified Extended (Mercor held-out)Code66.055.3−10.7
Snorkel SWE-bench CLICode16.66.1−10.5
BioSecBench-Surveillance V1 Clean RerunsScience58.548.0−10.5
LMArena Agent Arena Steering BurdenAgents18.6-2.3−10.4
Epoch Capabilities Index (ECI)Indexes169156−10.2
HealthBenchHealth & Medicine64.354.2−10.1
Artificial Analysis Coding Agent Index v1.5Indexes66.056.3−9.7
BenchBench-CloningScience20.510.9−9.6
Ruby LLM Benchmarks — Program FixerCode77.568.0−9.6
DAYJOB: FinanceFinance & Accounting23.914.5−9.4
CVDP (517-problem set)Code64.154.9−9.2
SpeciEval (Refusal-Aware Retries, Historical)Safety87.296.4+9.2
SpeciEval (Pooled English Runs)Safety87.296.4+9.2
HLAB-Vals-1.1Legal9.218.3+9.2
SciCode 1.0.1 (Artificial Analysis)Code66.957.8−9.1
Excel Modeling BenchmarkFinance & Accounting75.967.0−8.9
CloudOpsBench v1.0Agents47.856.5+8.7
PubMedQA (SecureBio)Health & Medicine87.979.2−8.7
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal10.819.5+8.7
AutoResearchExam (hidden-test AUARC)Agents0.60.5−8.7
SAFE-Teen 106 worst branch: Health Advice (under-18 prompt)Safety27.719.2+8.5
Snorkel Agentic Coding 2.0Code40.431.9−8.5
MMMU ProMultimodal86.678.2−8.4
GPQA-Bio (SecureBio)Science90.982.5−8.4
DataCenterBench Tier 1Agents21.713.3−8.3
ARC-AGI-1 Semi-PrivateGeneral QA98.590.2−8.3
HANDBOOK.md Agents v1.1Professional Work52.344.0−8.3
GDPxProfessional Work12.03.9−8.2
GDP.xlsxProfessional Work30.322.3−8.0
LMArena Agent Arena Confirmed SuccessAgents20.75.1−7.8
SnorkelUnderwrite 2.0Finance & Accounting22.730.5+7.8
RedlineBenchLegal42.049.5+7.5
WMDP-Bio (SecureBio)Science96.188.8−7.3
AA-Omniscience IndexIndexes46.432.0−7.2
SAFE-Teen 106: Urgent Safety (under-18 prompt)Safety3.110.1−6.9
AA-Briefcase v1.1 EloAgents17611636−6.9
SAFE-Teen Urgent Safety (under-18 prompt)Safety3.310.1−6.9
CharXiv Extended (Mercor held-out)Multimodal84.778.2−6.5
Ramp Accounting BenchFinance & Accounting50.544.1−6.4
ATLAS-Finance v1.0Finance & Accounting11.45.0−6.4
Finance Agent v2Finance & Accounting58.652.3−6.3
Vending-Bench Arena: Round 13Games80947930−6.3
ChessBench CoherenceGames80.186.3+6.2
EnterpriseBench: CoreCraftProfessional Work74.468.2−6.2
LegalReasonBenchLegal46.540.4−6.1
AA-LCR Extended (Mercor held-out)Search74.268.2−6.0
FrontierCode 1.1 Extended - ScoreCode65.359.4−5.9
DualEntry Accounting AI BenchmarkFinance & Accounting82.288.1+5.9
ChessBench AccuracyGames79.374.0−5.7
GDP.pdf (Artificial Analysis, All-pass)Professional Work28.823.2−5.6
FrontierCode 1.1 Extended - Pass RateCode70.865.2−5.6
SAFE-Teen 106: Health Advice (under-18 prompt)Safety17.312.0+5.3
GDPval-AA v2.1Professional Work18361717−5.2
SAFE-Teen 106 worst branch: Health Advice (no prompt)Safety44.339.2+5.1
SAGEMath45.840.8−5.0
MetagenomicsBench V0Science43.748.7+5.0
MetagenomicsBench V1Science43.748.7+5.0
WellnessBenchSafety53.048.0−5.0
SpatialBench-Long V1 Clean RerunsScience19.624.4+4.8
AA-LCR v1.1Search84.780.0−4.7
MMLU ProGeneral QA93.488.8−4.6
SM BenchSafety81.886.3+4.6
SAFE-Teen Health Advice (under-18 prompt)Safety15.110.6+4.5
CharXiv (descriptive and reasoning combined)Multimodal96.492.0−4.4
Excel Modeling Benchmark v1.1Finance & Accounting75.871.5−4.3
Snorkel Digital Work IndexIndexes23.619.3−4.3
DeepResearch Bench II (Mercor 130-task set)Search53.349.0−4.3
VariantBench V1 Clean RerunsScience49.745.5−4.2
DAYJOB: TaxFinance & Accounting9.25.0−4.2
Immigration BenchLegal72.168.0−4.1
MissionReportProfessional Work63.559.5−4.0
Veracode GenAI Code Security (Fall 2026)Cybersecurity73.069.0−4.0
LMArena Agent Arena Bash RecoveryAgents15.07.2−3.9
ITBench-AAAgents38.242.1+3.9
SWE-TogetherCode68.865.1−3.7
Political Science LLM BenchmarkProfessional Work70.266.6−3.6
DeepSWE v1.1Code74.271.0−3.2
DAXBench v2Code95.792.6−3.2
WCB (SecureBio)Science54.557.6+3.1
Biopharma Bench V0.1Health & Medicine67.564.4−3.1
Odysseys2Agents71.268.2−3.0
FrontierMath ErdősMath2.90.0−2.9
DTBenchGeneral QA98.996.0−2.9
LMArena Agent Arena Bash ReliabilityAgents15.19.3−2.9
SAFE-Teen 106 worst branch: Relationships (under-18 prompt)Safety2.45.1−2.7
AdvancedIFChat & Writing82.880.2−2.6
SAGE v1.1Math91.388.7−2.6
SAFE-Teen Urgent Safety (no prompt)Safety13.115.7−2.5
DAYJOB: LawLegal25.127.6+2.5
SAFE-Teen 106: Urgent Safety (no prompt)Safety12.715.1−2.4
Hermes BenchAgents76.774.4−2.4
ScBench-Long V1 Clean RerunsScience13.611.4−2.3
Visa Agentic Commerce BenchmarkAgents62.260.0−2.2
BioSimBench 1.0Science14.512.3−2.2
Olam Social Poker EloGames15061508+2.2
Financial Coaching BenchFinance & Accounting76.274.1−2.1
MedScribeHealth & Medicine91.489.4−2.1
ScBench V1 Clean RerunsScience56.458.5+2.1
RCM-Bench-1Health & Medicine90.188.2−2.0
Creative Writing v3 EloChat & Writing20512012−2.0
OTIS Mock AIME 2024-2025Math100.098.1−1.9
SWE Atlas - Codebase QnACode48.947.0−1.9
MMLU-Bio (SecureBio)Science92.790.9−1.8
dfbench Detection RecallCybersecurity54.956.6+1.7
SnorkelManufacturingProfessional Work10.412.1+1.7
SnorkelFinance 2.0Finance & Accounting18.620.2+1.6
Longform Creative WritingChat & Writing84.683.0−1.6
ForecastBench Preliminary Dataset Brier IndexForecasting62.260.7−1.5
Terminal-Bench 2.1Code83.585.0+1.5
SnorkelLegalLegal31.432.8+1.4
AA-AnalystAgent (pass^5)Professional Work56.355.0−1.2
SnorkelRevOpsProfessional Work12.511.3−1.2
ComplexConstraintsChat & Writing49.948.9−1.0
CWE-bench v1 Programmatic Pass@1Cybersecurity67.068.0+1.0
Eval Connections One-Shot with TrapsGames86.085.0−1.0
HealthBench (length-adjusted, Artificial Analysis)Health & Medicine57.056.0−1.0
Snorkel WorkplaceAgentsProfessional Work14.415.2+0.8
ObviousBenchGeneral QA100.099.3−0.7
SWE-Marathon v1.1Code46.345.6−0.6
SAFE-Teen Relationships (under-18 prompt)Safety0.51.1−0.6
AML BenchFinance & Accounting66.465.8−0.6
SAFE-Teen 106: Relationships (under-18 prompt)Safety0.41.0−0.6
Arena Alignment IndexSafety83.282.7−0.4
APEX-1 (Mercor, mean score)Professional Work65.865.4−0.4
MedCodeHealth & Medicine49.849.6−0.2
Ivo-micro1 ContractBenchLegal58.158.3+0.2
LMArena Agent Arena Tool HallucinationAgents0.71.0+0.2
Creative Writing v3 Rubric ScoreChat & Writing16.015.9−0.2
Access to Justice BenchmarkLegal73.873.7−0.1
AIME 2025Math100.0100.00.0
Dual-Arm Sim (task completion)Agents0.00.00.0
Harvey's Legal Agent Benchmark - Arbitration International Dispute ResolutionLegal0.00.00.0
Harvey's Legal Agent Benchmark - Banking FinanceLegal0.00.00.0
Harvey's Legal Agent Benchmark - Bankruptcy RestructuringLegal0.00.00.0
Harvey's Legal Agent Benchmark - Corporate GovernanceLegal0.00.00.0
Harvey's Legal Agent Benchmark - Data Privacy CybersecurityLegal0.00.00.0
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 - Energy Natural ResourcesLegal20.020.00.0
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 - International Trade SanctionsLegal0.00.00.0
Harvey's Legal Agent Benchmark - Real EstateLegal0.00.00.0
Harvey's Legal Agent Benchmark - Structured Finance SecuritizationLegal0.00.00.0
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
SpatialBench V1 Clean RerunsScience69.669.60.0

Claude Opus 5.5vsGrok 4.7normalised score marginsData Nov 15, 2026Data: Vals AI, Mercor, Artificial Analysis and 94 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…