Atlas

Compare Models

Claude Opus 5.5 vs GPT-5.6 Sol

Claude Opus 5.5 leads by 4.9 capability points (≈95% interval +2.4 to +7.3): a clear lead over GPT-5.6 Sol · P(higher) ≈ 100%

Wins 180 of 244 shared benchmarks (39 lost, 25 tied)

Capability vs. relative cost
Claude Opus 5.5GPT-5.6 SolClosed 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-5.6 Solcapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins180 – 39 · 244 shared, 25 tied

Comparing matching setups only (138 benchmarks excluded) · suites counted once
← Claude Opus 5.5GPT-5.6 Sol →

Claude Opus 5.5vsGPT-5.6 Solwins by categoryData Nov 15, 2026Data: Vals AI, Artificial Analysis, benchmarks.bio and 84 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is GPT-5.6 Sol minus Claude Opus 5.5.

KernelBench Mega Normalized (RTX PRO 6000)Code100.07.4−92.6
VCT-Refusal (CAIS)Safety90.411.3−79.1
HPCT Refusal (CAIS)Safety100.024.0−76.0
TasteVal (compute multiplier)Science2.30.6−75.3
YC-BenchGames2873725728001−74.9
CheatBenchSafety7.274.9−67.6
Time Horizon Index: KSPAgents91.323.8−67.5
LisanBench Path LengthGames240294699−60.1
MirrorCode (ML, +Private, 2L)Code78.920.0−58.9
LisanBench Difficulty-WeightedGames59631403−48.8
Medical Long Context Reasoning (MLCR-AA)Health & Medicine66.726.1−40.6
Terminal-Bench-Science 0.1Science61.922.4−39.5
AutomationBench 1.0.6 (Zapier)Agents42.53.8−38.7
BioSecBench-Function V1 Clean RerunsScience73.635.5−38.1
Terminal-Bench Science 0.1 (Vals)Science48.612.9−35.7
Search ArenaSearch12131258+35.7
TxBench Preclinical Pharmacology V1 Clean RerunsScience26.361.3+35.0
Terminal-Bench 4.0 (Artificial Analysis)Code59.624.7−34.8
TxBench Oligo Discovery V1 Clean RerunsScience1.435.0+33.6
Humanity's Last Exam RMS Calibration Error (CAIS)General QA13.846.7−32.9
BrokenArXiv 06/2026Math100.067.3−32.7
Olam Social Poker EloGames15061534+30.4
FrontierSWE V2 Worst@5Code52.022.2−29.8
BrowseComp Extended (Mercor held-out)Search60.532.3−28.2
EnterpriseBench: CoreCraftProfessional Work74.446.2−28.2
LMArena Agent Arena Net Improvement (Steering Burden)Agents19.54.6−28.1
Terminal-Bench 4.0Code64.837.3−27.5
FrontierSWE V2 Best@5Code71.544.2−27.3
BioTIER RefusalSafety98.771.9−26.9
HiL-Bench (Human-in-Loop Benchmark)Agents59.032.3−26.7
EBR-bench v4 (card ban, single agent)Games71.444.8−26.7
Furniture Assembly BenchmarkMultimodal83.356.7−26.7
Gray Swan IPI Attack Success (k=15)Safety1.027.0−26.0
TxBench Antibody Discovery V1 Clean RerunsScience6.831.4+24.6
TextQuests Average Cumulative HarmSafety24.231.4−24.4
Humanity's Last Exam DiamondGeneral QA55.031.2−23.8
Learn2Play BenchGames80.157.0−23.1
DAYJOB: HealthcareHealth & Medicine24.71.6−23.1
ScBench-Long V1 Clean RerunsScience6.829.6+22.7
VoxelBench Text RatingGames26132297−22.5
HardGeoBench Avg DistanceMultimodal6551381−22.4
Concept Synth INDUCTION Challenge100 Round 1General QA65.043.0−22.0
HardGeoBench Country AccuracyMultimodal68.847.5−21.2
Drone-BenchAgents97.376.1−21.2
SM BenchSafety81.861.1−20.7
τ³-Banking (Fireworks, banking_knowledge)Agents56.035.7−20.4
HardGeoBench <=100miMultimodal61.341.3−20.0
Harvey's Legal Agent Benchmark - Healthcare Life SciencesLegal20.00.0−20.0
BioSecBench-RespiratoryVirus-Surveillance V1 (110 evaluations)Science25.45.8−19.7
APEX-SWECode67.648.0−19.6
FrontierCUA pass@1Agents36.016.4−19.6
ProgramBench (Mercor 100-task subset)Code20.01.0−19.0
CrossSearch (pass^3)Search57.038.0−19.0
DAYJOB: FinanceFinance & Accounting23.95.5−18.4
LAB-Bench LitQA2 (SecureBio)Science71.152.8−18.3
ITBench-AAAgents38.256.2+18.0
ActiveVisionMultimodal41.223.5−17.7
AA-Briefcase v1.1 EloAgents18011479−17.7
Blueprint-Bench 2Multimodal0.50.3−17.6
SpatialBench-Long V1 Clean RerunsScience20.237.5+17.3
ProofBench v1.1Math100.083.0−17.0
LMArena Agent Arena Praise vs ComplaintAgents40.36.3−17.0
MindCube TinyMultimodal93.176.8−16.2
CursorBench 4.0Code57.841.7−16.1
Agent Red Teaming k=100 (CAIS)Safety24.440.3−15.9
Machiavelli Deception (CAIS)Safety77.582.7−15.8
VariantBench V1 Clean RerunsScience50.635.3−15.2
BullshitBench v2 LeaderboardSafety63.048.0−15.0
GameBench 2Games94.880.1−14.7
Humanity's Sixth SenseMultimodal44.630.0−14.6
LLM Creative Story WritingChat & Writing3.72.5−14.1
Concept Synth INDUCTION Challenge64 Round 1General QA71.957.8−14.1
Toloka Arena (v3)Agents75.061.0−14.0
VIALSMultimodal40.226.5−13.7
EpiBench V1 Clean RerunsScience39.025.8−13.2
Mystery Game PuzzlesGames71.058.0−13.0
ZeroBench Main Mean pass@1 (five samples)Multimodal35.022.0−13.0
HackerRank-BenchCode51.238.4−12.8
BenchCAD - geometric overlapMultimodal96.283.4−12.8
Hedge-BenchFinance & Accounting44.231.5−12.7
PCT Political Inconsistency (CAIS)Safety29.141.7−12.6
Conceptual Reasoning IndexIndexes77.968.7−12.6
ACCoRDGeneral QA88.475.9−12.5
HardGeoBench <=10miMultimodal42.530.0−12.5
AA-Omniscience IndexIndexes46.422.0−12.2
LAB-Bench ProtocolQA (SecureBio)Science94.081.9−12.1
FrontierMath Tier 4 (v2) - PrivateMath95.082.9−12.1
CadQueryEval (Corrected Geometry)Code100.088.0−12.0
TaxCalcBench TY25Finance & Accounting42.030.0−12.0
ZeroBench Main Questions pass^5Multimodal25.013.0−12.0
ALE-BenchCode27482410−12.0
SpeciEval (Refusal-Aware Retries, Historical)Safety87.299.0+11.8
SpeciEval (Pooled English Runs)Safety87.299.0+11.8
SciCode (main problems)Science27.316.2−11.1
VCT v2 (SecureBio)Science60.549.4−11.1
SWE-TogetherCode68.857.8−11.0
GDPval-AA v2.1Professional Work18661619−10.9
OncologyBench Strict (25 tasks, October 9 rubric)Health & Medicine29.318.7−10.7
APEX-AccountingFinance & Accounting61.851.5−10.3
LiveBench 2.0Agents86.576.3−10.3
Humanity's Last Exam (Text-Only)General QA57.547.3−10.2
Terminal-Bench 2.1 Extended (Mercor held-out)Code47.537.4−10.1
OpenDS-WorldSearch38.028.0−10.0
WCB (SecureBio)Science54.564.4+9.9
HieroglyphBenchMultimodal37.427.6−9.8
LMCAGeneral QA68.258.5−9.7
Humanity's Last Exam (with tools)General QA67.758.0−9.7
Agents' Last Exam Overall ScoreAgents63.253.6−9.6
AutoResearchExam (hidden-test AUARC)Agents0.60.5−9.6
Roboflow Vision Evals (October 2026, five tasks)Multimodal82.072.4−9.6
LibraryDesignBenchCode48.939.5−9.4
LMArena Agent Arena Confirmed SuccessAgents20.72.3−9.2
SciCode 1.0.1 (Artificial Analysis)Code66.957.8−9.1
Chartography (with tools)Multimodal89.079.9−9.1
ChessBench CoherenceGames80.189.1+9.0
LABBench2Science73.182.1+9.0
SpatialViz-BenchMultimodal89.780.7−9.0
HealthBenchHealth & Medicine64.355.3−9.0
ClockBench AccuracyMultimodal77.268.3−8.9
DeepsecBench — Revalidated V1Code26.735.4+8.8
AA-AnalystAgent (pass^5)Professional Work56.347.5−8.7
DeepsecBench — Revalidated V1 (233 issues)Code26.635.3+8.7
CVDP (517-problem set)Code64.155.4−8.7
BioSecBench-Structure V1 (58 evaluations)Science29.320.7−8.6
SUPERChemScience79.671.0−8.6
BioSecBench-Surveillance V1 Clean RerunsScience55.847.3−8.6
RedlineBenchLegal42.050.3+8.3
Nimble Web ContextSearch71.880.0+8.2
Chess Puzzles (task version 2.0)Games67.059.0−8.0
MetagenomicsBench V1Science39.331.3−8.0
MetagenomicsBench V0Science39.331.3−8.0
LLM Debate BenchmarkChat & Writing17011645−8.0
ChessBench EloGames13491218−7.8
ReactBench (39 tasks)Code54.446.7−7.7
MBCT (SecureBio)Science68.761.1−7.6
BuseyBench SVGMultimodal6.86.1−7.4
MaxxBenchCode76.469.1−7.3
LMArena Agent Arena Bash RecoveryAgents15.00.7−7.1
VCT (SecureBio)Science61.454.3−7.1
FORTRESS Average Risk Score (ARS)Safety13.220.1−7.0
PubMedQA (SecureBio)Health & Medicine87.981.1−6.8
ATLAS-Finance v1.0Finance & Accounting11.44.7−6.7
SAGEMath45.852.6+6.7
LMArena Agent Arena Steering BurdenAgents18.65.5−6.5
Roboflow Vision Evals (September 2026 protocol)Multimodal85.579.0−6.5
WMDP-Bio (SecureBio)Science96.189.7−6.4
GBENCH One-shot Coding GScoreGames73.967.7−6.2
MedScribeHealth & Medicine91.485.2−6.2
Terminal-Bench 2.1Code83.589.5+6.0
GeneBench-Pro Public Release SubsetScience30.024.0−6.0
MASKSafety89.495.4+6.0
MedCodeHealth & Medicine49.844.0−5.8
Longtail-SWECode81.575.7−5.8
Harvey LAB held-out set (Fireworks run)Legal6.30.6−5.7
GPQA-Bio (SecureBio)Science90.985.3−5.6
Extended NYT Connections 940 (Quadratic v1)Games88.593.8+5.3
TextQuests Game ProgressGames57.352.1−5.2
ArXivMath 06/2026Math83.388.5+5.2
Clocktower BenchGames15471583+5.2
Into the Unknown Direction Node Pass RateScience80.074.9−5.1
APEX-Agents 1.1 Corporate Law (68-task image-free slice)Legal85.480.6−4.8
RiemannBenchMath69.674.4+4.8
Harvey HLAB — Criteria Pass RateLegal75.880.5+4.8
dfbench Detection RecallCybersecurity54.959.6+4.7
MacroscopeBenchCode73.177.8+4.7
FrontierCode 1.1 Extended - ScoreCode65.360.6−4.7
SpatialBench V1 Clean RerunsScience72.868.1−4.6
SWE Atlas - Codebase QnACode48.953.5+4.6
Senior SWE-BenchCode39.334.7−4.6
CritPt v2 (agentic, avg@5)Science81.777.1−4.6
Epoch Capabilities Index (ECI)Indexes169163−4.6
ScBench V1 Clean RerunsScience55.459.8+4.4
IntPhys 2Multimodal68.363.9−4.3
BigFinanceBench private holdout (final-answer accuracy)Finance & Accounting48.244.2−4.0
WellnessBenchSafety53.057.0+4.0
MMMU ProMultimodal86.682.7−3.9
ERQAMultimodal75.171.2−3.9
Creative Writing v3 EloChat & Writing20511974−3.9
LMArena Text Arena EloChat & Writing15061485−3.7
SAB ResearchFinance & Accounting97.293.6−3.6
MCP Atlas April 2026 (all 1,000 tasks)Agents85.481.8−3.6
Excel Modeling BenchmarkFinance & Accounting75.972.3−3.6
ObviousBenchGeneral QA100.096.5−3.5
DTBenchGeneral QA98.995.7−3.2
CritPt (avg@5)Science31.728.6−3.1
ChessBench AccuracyGames79.376.4−3.1
Eval Connections One-Shot with TrapsGames86.083.0−3.0
Veracode GenAI Code Security (Fall 2026)Cybersecurity73.070.0−3.0
FrontierMath ErdősMath2.90.0−2.9
Longform Creative WritingChat & Writing84.681.7−2.9
AHK-EvalCode91.794.4+2.8
GlobeBenchGeneral QA26.323.7−2.7
Vectara Hallucination LeaderboardSafety9.912.4−2.5
Ruby LLM Benchmarks — Program FixerCode77.575.2−2.4
FrontierMath Tiers 1-3 (v2) - PrivateMath91.289.1−2.1
BrowseComp (Mercor 130-question subset)Search88.590.6+2.1
Short Form Bench v1.0Games23892410+2.1
DistressBenchSafety75.673.6−2.0
LMArena Agent Arena Bash ReliabilityAgents15.111.4−1.9
DeepResearch Bench II (Mercor 130-task set)Search53.351.5−1.8
Into the Unknown Integrity Node Pass RateScience90.488.7−1.6
MMLU-Bio (SecureBio)Science92.791.1−1.6
SAB RankingForecasting-0.1-0.1−1.6
Financial Coaching BenchFinance & Accounting76.274.9−1.3
Bio Capabilities Index (SecureBio)Indexes171169−1.2
LLM Chess EloGames15751550−1.0
Arena Alignment IndexSafety83.284.2+1.0
ARC-AGI-1 Semi-PrivateGeneral QA98.597.5−1.0
GDP.pdf (Artificial Analysis, All-pass)Professional Work28.827.8−1.0
Harvey LAB: In-House Contracting (held-out)Legal3.02.0−1.0
Design Creativity BenchCode51.051.3+1.0
Creative Writing v3 Rubric ScoreChat & Writing16.016.2+1.0
ARC-AGI-2General QA93.392.5−0.8
ForecastBench Preliminary Dataset Brier IndexForecasting62.261.4−0.8
AA-LCR v1.1Search84.784.0−0.7
Prophet Arena Arena Score (Fixed Context)Forecasting85.584.8−0.6
ComplexConstraintsChat & Writing49.950.5+0.6
DeepSWE v1.1Code72.372.7+0.4
LMArena Agent Arena Tool HallucinationAgents0.71.0+0.2
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 - 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 - 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
Into the UnknownScience0.00.00.0
OTIS Mock AIME 2024-2025Math100.0100.00.0

Claude Opus 5.5vsGPT-5.6 Solnormalised score marginsData Nov 15, 2026Data: Vals AI, Artificial Analysis, benchmarks.bio and 84 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…