Atlas

Compare Models

Claude Opus 5.5 vs GPT-5.5

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

Wins 146 of 198 shared benchmarks (26 lost, 26 tied)

Capability vs. relative cost
Claude Opus 5.5GPT-5.5Closed 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 5.5vsGPT-5.5capability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins146 – 26 · 198 shared, 26 tied

Comparing matching setups only (70 benchmarks excluded) · suites counted once
← Claude Opus 5.5GPT-5.5 →

Claude Opus 5.5vsGPT-5.5wins by categoryData Nov 15, 2026Data: Vals AI, Mercor, Center for AI Safety and 57 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is GPT-5.5 minus Claude Opus 5.5.

KernelBench Mega Normalized (RTX PRO 6000)Code100.012.2−87.8
Time Horizon Index: KSPAgents91.38.3−83.0
TasteVal (compute multiplier)Science2.30.5−80.6
VCT-Refusal (CAIS)Safety90.416.0−74.4
HPCT Refusal (CAIS)Safety100.029.0−71.0
MirrorCode (ML, +Private, 2L)Code78.910.0−68.9
YC-BenchGames28737251206045−58.2
VoxelBench Text RatingGames26131925−49.0
Medical Long Context Reasoning (MLCR-AA)Health & Medicine66.717.8−48.9
BioSecBench-Function V1 Clean RerunsScience73.628.3−45.3
BrokenArXiv 05/2026Math95.050.0−45.0
Terminal-Bench 4.0 (Artificial Analysis)Code59.614.6−44.9
LisanBench Path LengthGames240299845−44.1
Code2Video-AA EloCode15361000−43.5
BrowseComp Extended (Mercor held-out)Search60.519.0−41.5
AutomationBench 1.0.6 (Zapier)Agents42.51.1−41.4
Furniture Assembly BenchmarkMultimodal83.344.2−39.2
GameBench 2Games94.859.9−34.8
AA-Briefcase v1.1 EloAgents17611135−34.3
Drone-BenchAgents97.363.9−33.4
ClockBench AccuracyMultimodal77.246.1−31.1
APEX-SWECode67.636.6−31.0
ActiveVisionMultimodal41.210.6−30.6
BrokenArXiv 06/2026Math100.069.4−30.6
TxBench Oligo Discovery V1 Clean RerunsScience1.431.9+30.6
TxBench Preclinical Pharmacology V1 Clean RerunsScience26.355.3+29.0
LisanBench Difficulty-WeightedGames59633316−28.4
Humanity's Last Exam RMS Calibration Error (CAIS)General QA13.841.8−27.9
Humanity's Last Exam DiamondGeneral QA54.627.9−26.7
SWE-bench Verified Extended (Mercor held-out)Code66.039.3−26.7
SWE Atlas Codebase QnA Extended (Mercor held-out)Code43.317.3−26.0
BioTIER RefusalSafety98.773.4−25.4
LLM Debate BenchmarkChat & Writing17011528−24.7
CadQueryEval (Corrected Geometry)Code100.076.0−24.0
HardGeoBench Country AccuracyMultimodal68.845.0−23.7
TxBench Antibody Discovery V1 Clean RerunsScience6.829.9+23.1
EnterpriseBench: CoreCraftProfessional Work74.451.3−23.1
ALE-BenchCode27482108−22.7
FrontierMath Tier 4 (v2) - PrivateMath95.072.5−22.5
HardGeoBench <=100miMultimodal62.540.0−22.5
GeneBench-Pro Public Release SubsetScience30.08.0−22.0
LAB-Bench LitQA2 (SecureBio)Science71.149.8−21.3
GDPval Extended (Mercor held-out rubric score)Professional Work75.854.5−21.3
GDPval-AA v2.1Professional Work18361357−21.1
Design Arena Overall Code CategoriesCode13841257−21.1
EnigmaEvalGeneral QA57.837.2−20.6
VariantBench V1 Clean RerunsScience50.630.5−20.0
Harvey's Legal Agent Benchmark - Data Privacy CybersecurityLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Healthcare Life SciencesLegal20.00.0−20.0
ChessBench CoherenceGames80.1100.0+19.9
ExploitGymCybersecurity34.515.1−19.4
SimpleBenchGeneral QA88.469.0−19.4
SWE-bench VerifiedCode98.278.8−19.4
HiL-Bench (Human-in-Loop Benchmark)Agents59.039.7−19.3
ProgramBench (Mercor 100-task subset)Code20.01.0−19.0
TaxCalcBench TY25Finance & Accounting42.024.0−18.0
LiveBench 2.0Agents86.568.7−17.8
ZeroBench Main Mean pass@1 (five samples)Multimodal35.017.2−17.8
SWE-bench MultilingualCode95.577.7−17.8
MineBench — Bradley-Terry RatingGames21341846−17.4
Agent Red Teaming k=100 (CAIS)Safety24.441.5−17.1
HardGeoBench Avg DistanceMultimodal6551196−16.7
SM BenchSafety81.865.4−16.3
Olam Social Poker EloGames15061521+16.3
AURORA Multilingual Terminal-BenchCode76.560.3−16.2
LAB-Bench ProtocolQA (SecureBio)Science94.077.9−16.1
BullshitBench v2 LeaderboardSafety63.047.0−16.0
Roboflow Vision Evals (October 2026, five tasks)Multimodal82.066.0−16.0
SpatialViz-BenchMultimodal89.774.2−15.5
Humanity's Last Exam (with tools)General QA67.752.2−15.5
VCT v2 (SecureBio)Science60.545.3−15.2
Agents' Last Exam (ALE)Agents38.223.0−15.1
Blueprint-Bench 2Multimodal0.50.4−15.0
HardGeoBench <=10miMultimodal43.828.8−15.0
SciCode (main problems)Science27.312.3−15.0
Mystery Game PuzzlesGames71.056.0−15.0
ScBench-Long V1 Clean RerunsScience6.821.6+14.8
OncologyBench Strict (25 tasks, October 9 rubric)Health & Medicine29.314.7−14.7
HieroglyphBenchMultimodal37.422.8−14.6
LLM Creative Story WritingChat & Writing3.72.4−14.5
Conceptual Reasoning IndexIndexes77.967.3−14.5
RiemannBenchMath69.655.2−14.4
ZeroBench Main Questions pass^5Multimodal25.011.0−14.0
LMCAGeneral QA68.254.3−14.0
APEX-AccountingFinance & Accounting61.848.2−13.6
MindCube TinyMultimodal93.179.9−13.2
AA-Omniscience IndexIndexes46.420.5−13.0
SUPERChemScience79.666.8−12.8
Machiavelli Deception (CAIS)Safety77.581.7−12.8
Humanity's Last Exam (Text-Only)General QA57.545.0−12.5
SpatialBench-Long V1 Clean RerunsScience20.232.1+11.9
GBENCH One-shot Coding GScoreGames73.962.2−11.7
MBCT (SecureBio)Science68.757.1−11.6
Excel Modeling BenchmarkFinance & Accounting75.964.5−11.4
APEX-Agents 1.1Professional Work81.370.3−11.0
SpeciEval (Pooled English Runs)Safety87.298.2+11.0
SpeciEval (Refusal-Aware Retries, Historical)Safety87.298.2+11.0
PCT Political Inconsistency (CAIS)Safety29.140.0−10.9
BrowseComp (Mercor 130-question subset)Search88.577.7−10.8
SciCode 1.0.1 (Artificial Analysis)Code66.956.1−10.8
Roboflow Vision Evals (September 2026 protocol)Multimodal85.574.8−10.7
EpiBench V1 Clean RerunsScience39.028.3−10.7
Creative Writing v3 EloChat & Writing20511845−10.4
MedXpertQA MM Extended (Mercor held-out)Health & Medicine71.260.8−10.4
DeepSWE v1.1Code74.264.4−9.8
Senior SWE-BenchCode39.329.5−9.8
TextQuests Average Cumulative HarmSafety24.221.4+9.5
Terminal-Bench 2.1 Extended (Mercor held-out)Code47.538.0−9.5
Bio Capabilities Index (SecureBio)Indexes171158−9.4
SimpleQA Verified (Epoch anti-abstention)General QA72.263.0−9.2
VCT (SecureBio)Science61.452.3−9.1
MedXpertQA MM (200-question set)Health & Medicine85.776.7−9.0
Clocktower BenchGames15471486−8.8
BioSecBench-Surveillance V1 Clean RerunsScience55.847.1−8.7
SpatialBench V1 Clean RerunsScience72.864.1−8.7
BuseyBench SVGMultimodal6.86.0−8.5
TextQuests Game ProgressGames57.348.9−8.4
IntPhys 2Multimodal68.359.9−8.4
ARC-AGI-2General QA93.385.0−8.3
ACCoRDGeneral QA88.480.5−7.9
DeepResearch Bench II (Mercor 130-task set)Search53.345.4−7.9
Extended NYT Connections 940 (Quadratic v1)Games88.596.2+7.7
ITBench-AAAgents38.245.8+7.6
DAXBench v2Code95.788.3−7.4
PubMedQA (SecureBio)Health & Medicine87.980.5−7.4
MMMU-Pro Extended (Mercor held-out)Multimodal51.043.8−7.2
Epoch Capabilities Index (ECI)Indexes169160−7.0
Longform Creative WritingChat & Writing84.678.2−6.4
CritPt (avg@5)Science31.725.4−6.3
AA-AnalystAgent (pass^5)Professional Work56.350.0−6.2
GDP.pdf (Artificial Analysis, All-pass)Professional Work28.822.6−6.2
FrontierMath Tiers 1-3 (v2) - PrivateMath91.285.3−6.0
SAGEMath45.851.5+5.7
DeepsecBench — Revalidated V1Code26.721.1−5.6
AHK-EvalCode91.797.2+5.6
DeepsecBench — Revalidated V1 (233 issues)Code26.621.0−5.5
MMMU ProMultimodal86.681.2−5.4
WCB (SecureBio)Science54.559.7+5.2
ERQAMultimodal75.170.5−4.6
MedScribeHealth & Medicine91.486.9−4.6
MMLU ProGeneral QA93.488.9−4.5
LMArena Text Arena EloChat & Writing15061481−4.3
Harvey HLAB — Criteria Pass RateLegal75.880.0+4.2
ArXivMath 05/2026Math81.777.5−4.2
Chess Puzzles (task version 2.0)Games67.063.0−4.0
AA-LCR Extended (Mercor held-out)Search74.270.5−3.7
GDPval (Mercor 213-task rubric score)Professional Work88.384.6−3.7
SWE Atlas - Codebase QnACode48.945.2−3.7
ARC-AGI-1 Semi-PrivateGeneral QA98.595.0−3.5
GPQA-Bio (SecureBio)Science90.987.4−3.5
FrontierMath ErdősMath2.90.0−2.9
DTBenchGeneral QA98.996.0−2.9
LMArena Agent Arena Bash RecoveryAgents15.09.5−2.7
APEX-1 (Mercor, mean score)Professional Work65.868.4+2.6
RedlineBenchLegal42.044.1+2.1
CharXiv Extended (Mercor held-out)Multimodal84.782.7−2.0
CharXiv (descriptive and reasoning combined)Multimodal96.494.6−1.8
LLM Chess EloGames15751532−1.8
ScBench V1 Clean RerunsScience55.457.1+1.7
AdvancedIFChat & Writing82.881.3−1.5
GlobeBenchGeneral QA26.325.2−1.2
ForecastBench Preliminary Dataset Brier IndexForecasting62.263.3+1.1
Terminal-Bench 2.1Code83.584.3+0.8
MASKSafety89.490.1+0.7
MedCodeHealth & Medicine49.849.1−0.7
Vectara Hallucination LeaderboardSafety9.99.3+0.6
ChessBench AccuracyGames79.378.8−0.5
Creative Writing v3 Rubric ScoreChat & Writing16.015.9−0.5
ComplexConstraintsChat & Writing49.949.5−0.4
AA-LCR v1.1Search84.784.3−0.3
ArXivMath 06/2026Math83.383.6+0.3
Ruby LLM Benchmarks — Program FixerCode77.577.60.0
AIME 2025Math100.0100.00.0
ChessBench EloGames134913490.0
Harvey's Legal Agent Benchmark - Antitrust CompetitionLegal0.00.00.0
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 - Capital MarketsLegal0.00.00.0
Harvey's Legal Agent Benchmark - Corporate GovernanceLegal0.00.00.0
Harvey's Legal Agent Benchmark - Corporate M&ALegal0.00.00.0
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 - Funds Asset ManagementLegal0.00.00.0
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 - Litigation Dispute ResolutionLegal0.00.00.0
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
ObviousBenchGeneral QA100.0100.00.0
OTIS Mock AIME 2024-2025Math100.0100.00.0

Claude Opus 5.5vsGPT-5.5normalised score marginsData Nov 15, 2026Data: Vals AI, Mercor, Center for AI Safety 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…