Atlas

Compare Models

Claude Opus 5 vs GPT-5.6 Sol

Claude Opus 5 leads by 1.2 capability points (≈95% interval −1.1 to +3.5): statistically tied · P(higher) ≈ 85%

Wins 215 of 346 shared benchmarks (103 lost, 28 tied)

Capability vs. relative cost
Claude Opus 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 5vsGPT-5.6 Solcapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins215 – 103 · 346 shared, 28 tied

Comparing matching setups only (169 benchmarks excluded) · suites counted once
← Claude Opus 5GPT-5.6 Sol →

Claude Opus 5vsGPT-5.6 Solwins by categoryData Nov 15, 2026Data: Vals AI, benchmarks.bio, Artificial Analysis and 100 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.

CyberBench PoCCybersecurity0.093.2+93.2
KernelBench Mega Normalized (H100)Code100.018.6−81.4
HPCT Refusal (CAIS)Safety100.024.0−76.0
DecepEval Long-horizon (Neutral)Safety58.31.5+56.8
DuetBench–Diagnosis EloHealth & Medicine1035990−53.2
CyberBench OverallCybersecurity40.788.1+47.5
Machiavelli Deception (CAIS)Safety96.582.7+41.7
YC-BenchGames1911592728001−41.3
LisanBench Difficulty-WeightedGames50681403−39.3
DecepEval Long-horizon (Induced)Safety83.347.8+35.5
KernelBench Hard Normalized (H100)Code82.848.6−34.1
LisanBench Path LengthGames154284699−33.3
Medical Long Context Reasoning (MLCR-AA)Health & Medicine59.426.1−33.3
DecepEval Coding (Induced)Safety64.331.7+32.7
TasteVal (compute multiplier)Science1.30.6−30.6
TextQuests Average Cumulative HarmSafety22.631.4−29.7
CheatBenchSafety45.374.9−29.5
Into the Unknown Integrity Node Pass RateScience59.588.7+29.2
BioTIER RefusalSafety100.071.9−28.1
ARC-AGI-3 (Semi-Private)General QA30.22.1−28.0
Document ArenaProfessional Work15091480−25.6
ChessBench CoherenceGames63.889.1+25.3
BullshitBench v2 LeaderboardSafety73.048.0−25.0
Concept Synth INDUCTION Challenge64 Round 1General QA32.857.8+25.0
Learn2Play BenchGames81.857.0−24.8
SM BenchSafety85.861.1−24.7
HiL-Bench (Human-in-Loop Benchmark)Agents57.032.3−24.7
BrokenArXiv 06/2026Math90.767.3−23.5
ChessBench EloGames8261218+23.3
MetagenomicsBench V0Science54.031.3−22.7
MetagenomicsBench V1Science54.031.3−22.7
EnterpriseBench: CoreCraftProfessional Work68.746.2−22.5
OncologyBench Strict (25 tasks, October 9 rubric)Health & Medicine40.018.7−21.3
SciUniverse Level 1Science30.59.4−21.1
Harvey's Legal Agent Benchmark - Corporate M&ALegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Healthcare Life SciencesLegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Structured Finance SecuritizationLegal20.00.0−20.0
FrontierSWE V2Code52.032.2−19.8
MedCodeHealth & Medicine63.644.0−19.6
Concept Synth INDUCTION Challenge100 Round 1General QA24.043.0+19.0
DecepEval Coding (Neutral)Safety20.41.6+18.8
NL2Repo - release-reported repository generation scoreCode75.356.8−18.5
VCT-Refusal (CAIS)Safety29.411.3−18.1
TxBench Antibody Discovery V0Science51.433.8−17.6
Olam Social Poker EloGames15181534+17.4
Terminal-Bench 4.0Code53.937.3−16.6
Chess Puzzles (task version 2.0)Games43.059.0+16.0
ProofBench v1.1Math99.083.0−16.0
APEX-SWECode63.748.0−15.7
VariantBench V1 Clean RerunsScience50.635.3−15.2
Search ArenaSearch12391258+15.1
ORCA-bench Hard RCA (Fireworks 324-task subset)Agents44.329.5−14.8
SWE-Bench Pro (Public Dataset)Code79.264.6−14.6
Humanity's Last Exam DiamondGeneral QA44.830.4−14.4
LLM Creative Story WritingChat & Writing3.72.5−14.3
dfbench Detection RecallCybersecurity45.459.6+14.2
ProgramBench - almost solved rollout percentageCode37.023.0−14.0
ComplexConstraintsChat & Writing37.350.5+13.2
HieroglyphBenchMultimodal14.627.6+13.0
ActiveVisionMultimodal10.623.5+12.9
HealthBenchHealth & Medicine67.155.3−11.8
BrowseComp Extended (Mercor held-out)Search44.032.3−11.7
ExploitGym - release-reported percentage, cohort unspecifiedCybersecurity22.133.7+11.6
LLM Debate BenchmarkChat & Writing17241645−11.3
SpeciEval (Pooled English Runs)Safety87.999.0+11.1
SpeciEval (Refusal-Aware Retries, Historical)Safety87.999.0+11.1
Humanity's Last ExamGeneral QA56.445.5−10.9
LLM Chess EloGames12851550+10.8
CorpFin v2 - Max Fitting ContextFinance & Accounting76.065.2−10.8
LMArena Agent Arena Net ImprovementAgents9.56.3−10.8
Drone-BenchAgents86.976.1−10.8
CyberGymCybersecurity94.283.6−10.6
TxBench Antibody Discovery V1 Clean RerunsScience41.931.4−10.5
Public Benefits Bench v1.1Professional Work76.966.5−10.4
BioSecBench-RefusalSafety29.439.5+10.1
Antidote: Everyday EditionSafety10551027−10.0
ChessBench — Historical Output RepairGames32.742.7+10.0
TaxCalcBench TY25Finance & Accounting20.030.0+10.0
AA-Briefcase EloAgents16451475−9.9
SimpleQA VerifiedGeneral QA59.969.7+9.8
FrontierMath Tier 4 (v2) - PrivateMath73.282.9+9.8
ITSMBenchAgents50.640.9−9.7
SpatialBench-Long V1 Clean RerunsScience28.037.5+9.5
Vending-Bench 2Games111829619−9.5
Terminal-Bench 4.0 (Artificial Analysis)Code49.039.9−9.1
Eval Connections One-Shot with TrapsGames74.083.0+9.0
Terminal-Bench 3 - release-reported accuracyCode43.334.4−8.9
HANDBOOK.mdProfessional Work32.323.5−8.8
BioSecBench-RespiratoryVirus-Surveillance V1 (110 evaluations)Science14.65.8−8.8
Tax Agent BenchFinance & Accounting76.768.0−8.7
Arena Alignment IndexSafety75.684.2+8.6
CorpFin v2 - Shared Max ContextFinance & Accounting74.966.4−8.5
CorpFin v2 (Vals Index component)Finance & Accounting74.966.4−8.5
CorpFin v2 (Vals Multimodal Index component)Finance & Accounting74.966.4−8.5
Creative Writing v3 EloChat & Writing21351974−8.2
BioMysteryBench (Vals Terminus 2)Science79.371.1−8.1
Conceptual Reasoning IndexIndexes74.768.7−8.1
Into the Unknown Direction Node Pass RateScience66.874.9+8.1
CadQueryEval (Corrected Geometry)Code80.088.0+8.0
OSWorld 2.0 500-step Partial ScoreAgents70.662.6−8.0
ACCoRDGeneral QA83.875.9−7.9
Vibe Code Bench v1.1Code88.480.5−7.9
MacroscopeBenchCode69.977.8+7.9
SpatialBench V0 VerifiedScience66.774.5+7.8
GlobeBenchGeneral QA15.923.7+7.7
TxBench Oligo Discovery V0Science42.234.5−7.7
ATLAS-Finance v1.0Finance & Accounting12.34.7−7.7
WitnessBench v1.0Games30.623.0−7.6
ScBench-Long V0 VerifiedScience36.443.9+7.6
AA-Omniscience IndexIndexes37.122.0−7.6
ALE-BenchCode26222410−7.5
MindCube TinyMultimodal84.376.8−7.5
VariantBenchScience49.442.1−7.3
Legal Research BenchLegal55.348.1−7.2
Harvey HLAB — Criteria Pass RateLegal87.780.5−7.2
Terminal-Bench Science 0.1 (Vals)Science27.120.0−7.1
EBR-bench v2Games45.738.6−7.1
CorpFin v2 - Exact PagesFinance & Accounting68.661.5−7.1
DecepEval Tool-use (Neutral)Safety9.32.2+7.1
AutomationBench (Overall)Agents25.118.1−7.0
Wiz Cyber Model Arena v3.9Cybersecurity71.264.2−7.0
BioSecBench-Function V1 Clean RerunsScience42.535.5−6.9
EpiBenchScience35.242.1+6.9
IOI (Vals v2)Code84.391.2+6.8
CVDP (517-problem set)Code62.255.4−6.8
DAYJOB: HealthcareHealth & Medicine8.41.6−6.8
Harvey LAB held-out set (Fireworks run)Legal7.40.6−6.8
Bio Capabilities Index (SecureBio)Indexes178169−6.7
GDP.pdfProfessional Work24.030.7+6.7
ATLAS-Finance v1.1Finance & Accounting10.74.0−6.7
OSWorld 2 (Aug 8) Offline, 500-step Binary AccuracyAgents34.728.1−6.6
AutoResearchExam (hidden-test AUARC)Agents0.60.5−6.6
HardGeoBench Avg DistanceMultimodal12181429−6.5
BenchCAD - geometric overlapMultimodal89.983.4−6.5
Vals Science IndexIndexes47.941.5−6.4
UnityBench-Mini3DCode68.862.3−6.4
LiveCodeBench (Vals AI public dataset)Code89.082.6−6.4
LMArena Agent Arena Bash RecoveryAgents13.60.7−6.4
RiemannBenchMath68.074.4+6.4
A-Fantasia (Valid Attempts with Retries)Games29.736.0−6.3
RedlineBenchLegal44.050.3+6.3
GDPval-AA v2 (normalized)Professional Work67.961.6−6.3
AA-AnalystAgent (pass^5)Professional Work53.847.5−6.2
HardGeoBench Country AccuracyMultimodal53.847.5−6.2
Terminal-Bench-Science 0.1Science28.622.4−6.2
GDPval-AA v2Professional Work17081585−6.1
SWE-Clar (implementation)Code63.057.0−6.0
SWE-TogetherCode63.857.8−6.0
LMCAGeneral QA64.558.5−6.0
PWNBench v0.1Cybersecurity60.054.0−6.0
SWE-Clar (clarification)Code84.078.1−5.9
LiveBench 2.0Agents82.176.3−5.8
DAYJOB: FinanceFinance & Accounting11.35.5−5.8
ChessBench AccuracyGames71.076.4+5.8
MedScribeHealth & Medicine91.085.2−5.8
CritPt v2 (agentic, avg@5)Science71.477.1+5.7
Toolathlon-VerifiedAgents80.674.9−5.7
MaxxBenchCode63.569.1+5.6
OSWorld 2.0 - reported partial reward, patch and step limit unspecifiedAgents68.362.7−5.6
SimpleVQA with toolsMultimodal70.364.7−5.6
Toloka Arena (v1)Agents52.947.6−5.3
BabyVision with toolsMultimodal94.188.9−5.2
BioSecBench-Structure V1 (58 evaluations)Science15.520.7+5.2
Ramp SWE-Bench (78 published tasks)Code88.583.3−5.1
BigFinanceBench private holdout (final-answer accuracy)Finance & Accounting49.244.2−5.0
HardGeoBench <=100miMultimodal46.341.3−5.0
IOICode91.786.7−5.0
PCT Political Inconsistency (CAIS)Safety36.741.7−5.0
Time Horizon Index: KSPAgents18.823.8+5.0
TxBench Oligo Discovery V1 Clean RerunsScience40.035.0−5.0
Co-Design BenchCode66.161.1−5.0
SpatialViz-BenchMultimodal75.880.7+4.9
CursorBench 4.0Code46.641.7−4.9
LMArena Agent Arena Net Improvement (Steering Burden)Agents7.24.6−4.9
GDPval-AA v2.1Professional Work17291619−4.9
ZeroBench Main Mean pass@1 (five samples)Multimodal17.222.0+4.8
MortgageTax (Vals all 1,258 documents)Finance & Accounting72.167.3−4.8
GameBench 2Games84.880.1−4.8
ABC-Bench liquid handling (SecureBio)Science94.098.7+4.7
BioSecBench-Function V0Science43.038.3−4.7
Ruby LLM Benchmarks — Program FixerCode79.875.2−4.7
DeepResearch Bench II (Mercor 130-task set)Search56.151.5−4.6
SCI-AGENT-EXAMScience64.259.6−4.6
Longform Creative WritingChat & Writing86.381.7−4.6
EnterpriseOps-Gym-AAAgents47.542.9−4.6
Code MigrationCode57.552.9−4.6
SpatialBench-Long V0 VerifiedScience31.836.4+4.5
WCB (SecureBio)Science68.864.4−4.4
BioSecBench-Surveillance V1 Clean RerunsScience51.547.3−4.2
Furniture Assembly BenchmarkMultimodal60.856.7−4.2
HackerRank-BenchCode42.638.4−4.1
OSWorld 2.0 - reported binary completion, patch and step limit unspecifiedAgents31.427.3−4.1
Chartography (with tools)Multimodal84.079.9−4.1
OSWorld 2 (Aug 8) Full, 500-step Binary AccuracyAgents31.427.3−4.1
MysteryMechanismScience37.433.3−4.1
CrossSearch (pass^3)Search34.038.0+4.0
LMArena Agent Arena Confirmed SuccessAgents10.22.3−3.9
Artificial Analysis Intelligence Index v4.3.2Indexes50.847.0−3.8
Odysseys2Agents58.362.0+3.7
SUPERChemScience74.771.0−3.7
Artificial Analysis Intelligence Index v4.3Indexes50.747.1−3.6
BankerToolBench v1.1Finance & Accounting32.629.0−3.6
ReactBench v1Code49.052.5+3.5
FrontierMath Tiers 1-3 (v2) - PrivateMath85.689.1+3.5
AutomationBench-AA 1.0.6Agents56.660.1+3.5
Bedside BenchHealth & Medicine94.691.1−3.5
ObviousBenchGeneral QA100.096.5−3.5
BALROGGames63.460.0−3.4
SWE-sweepCode1.34.7+3.4
DistressBenchSafety70.273.6+3.4
ProgramBench (Mercor 100-task subset)Code4.31.0−3.3
FORTRESS Average Risk Score (ARS)Safety16.820.1−3.3
Blueprint-Bench 2Multimodal0.30.3+3.2
APEX-Agents 1.1 Corporate Law (68-task image-free slice)Legal83.880.6−3.2
ScBench-LongScience41.338.1−3.2
SAGEMath49.452.6+3.1
NoCoT-Bench NCRI 15.2General QA115109−3.1
WeirdML (v2)Science91.888.8−3.0
DeepsecBenchCode32.635.6+3.0
APEX-AccountingFinance & Accounting54.551.5−3.0
GDP.pdf - all-pass rate reported by GoogleProfessional Work37.040.0+3.0
Realm Financial ReasoningFinance & Accounting47.444.4−3.0
DeepsecBench — Revalidated V1Code32.435.4+3.0
DeepsecBench — Revalidated V1 (233 issues)Code32.335.3+3.0
Agent Red Teaming k=100 (CAIS)Safety43.340.3+3.0
AutoformBenchMath61.958.9−2.9
LMArena Agent Arena Praise vs ComplaintAgents12.26.3−2.9
VCT (SecureBio)Science51.454.3+2.9
ClockBench AccuracyMultimodal65.568.3+2.8
CursorBench 3.2Code70.067.2−2.8
Hemingway-benchChat & Writing10471057+2.8
GPQA-Bio (SecureBio)Science82.785.3+2.6
SAB ResearchFinance & Accounting96.293.6−2.6
Hedge-BenchFinance & Accounting29.031.5+2.5
GPQA DiamondGeneral QA92.494.9+2.5
HardGeoBench <=10miMultimodal32.530.0−2.5
InteractiveBench (task pass)Agents33.330.8−2.5
DeepSWE v1.1Code73.270.7−2.4
HPCT (SecureBio)Science78.676.3−2.3
ScBench-Long V1 Clean RerunsScience27.329.6+2.3
Agents' Last Exam Overall ScoreAgents55.953.6−2.3
Political Science LLM BenchmarkProfessional Work71.469.2−2.2
DTBenchGeneral QA97.995.7−2.1
LABBench2Science84.282.1−2.1
VIALSMultimodal24.426.5+2.1
ARC-AGI-2General QA90.492.5+2.1
LMArena Agent Arena SteerabilityAgents10.76.5−2.1
AA-LCR v1.1Search82.084.0+2.0
GeneBench-Pro Public Release SubsetScience22.024.0+2.0
MASKSafety93.495.4+2.0
Shroombench (mushroom identification, top-1)Multimodal44.042.0−2.0
ZeroBench Main Questions pass^5Multimodal11.013.0+2.0
Agents' Last Exam - release-reported pass rate, subset unspecifiedAgents28.626.7−1.9
BuseyBench SVGMultimodal6.36.1−1.9
AutomationBench 1.0.6 (Zapier)Agents26.928.8+1.8
Bench to the Future 3 — Pooled ScoreForecasting0.10.1−1.7
CyberBench PatchCybersecurity81.483.1+1.7
TxBench Preclinical Pharmacology V1 Clean RerunsScience63.061.3−1.7
Bench to the Future 3 — September Panel — PooledForecasting0.10.1−1.7
Finance Agent v2 (Vals Index subset)Finance & Accounting56.655.0−1.6
Finance Agent v2 (Vals Multimodal Index component)Finance & Accounting56.655.0−1.6
ScBench V1 Clean RerunsScience58.359.8+1.5
CharXiv Reasoning (with tools)Multimodal89.387.8−1.5
DeepSearchQA (F1 score)Search94.693.1−1.5
Roboflow Vision Evals (October 2026, five tasks)Multimodal70.972.4+1.5
Terminal-Bench 2.1Code88.089.5+1.5
CRM-Agent-BenchAgents82.781.3−1.4
BrowseCompSearch90.892.2+1.4
LAB-Bench LitQA2 (SecureBio)Science51.452.8+1.4
MMMU ProMultimodal84.082.7−1.3
SciCode (Artificial Analysis avg@3)Science55.756.9+1.3
Artificial Analysis Agentic IndexIndexes55.354.0−1.3
TextQuests Game ProgressGames53.352.1−1.2
Excel Modeling BenchmarkFinance & Accounting73.672.3−1.2
LAB-Bench CloningScenarios (SecureBio)Science100.098.8−1.2
IntPhys 2Multimodal65.163.9−1.2
Epoch Capabilities Index (ECI)Indexes164163−1.2
MRCR v2 (8-needle, Context Arena) - 128K pointwiseGeneral QA91.392.4+1.1
OTIS Mock AIME 2024-2025Math98.9100.0+1.1
SciCode (main problems)Science17.316.2−1.1
Mystery Game PuzzlesGames59.058.0−1.0
AA-LCRSearch70.071.0+1.0
ProofBenchMath78.077.0−1.0
Toloka Arena (v3)Agents62.061.0−1.0
WordleBenchGames96.097.0+1.0
SAB RankingForecasting-0.1-0.1+1.0
EBR-bench v4 (card ban, single agent)Games45.744.8−1.0
MBCT (SecureBio)Science60.261.1+0.9
NoCoT-Bench NCKI (kspine v3)General QA119121+0.9
LMArena Text Arena EloChat & Writing14901485−0.8
LMArena Agent Arena Bash ReliabilityAgents9.811.4+0.8
LAB-Bench ProtocolQA (SecureBio)Science81.181.9+0.8
SWE-bench VerifiedCode97.096.2−0.8
Artificial Analysis Progress Index v1Indexes334331−0.8
DecepEval Tool-use (Induced)Safety36.036.8−0.8
Roboflow Vision Evals (September 2026 protocol)Multimodal78.379.0+0.7
Vision Arena OverallMultimodal12851288+0.7
EpiBench V1 Clean RerunsScience25.225.8+0.6
ForecastBench Preliminary Dataset Brier IndexForecasting62.061.4−0.6
Terminal-Bench 2.1 Extended (Mercor held-out)Code38.037.4−0.6
LMArena Agent Arena Steering BurdenAgents4.45.5+0.6
CritPt (avg@5)Science29.128.6−0.6
Extended NYT Connections 940 (Quadratic v1)Games94.393.8−0.5
ForecastBench Tournament Overall Brier IndexForecasting66.666.1−0.5
Humanity's Sixth SenseMultimodal30.530.0−0.5
GBENCH One-shot Coding GScoreGames68.167.7−0.5
Creative Writing v3 Rubric ScoreChat & Writing16.316.2−0.5
𝜏³-BankingAgents44.744.3−0.4
TaxEval v2Finance & Accounting75.174.8−0.4
Artificial Analysis Coding IndexIndexes78.078.3+0.4
MRCR v2 (8-needle, Context Arena) - 128K AUCGeneral QA97.597.1−0.3
ERQAMultimodal70.871.2+0.3
SpatialBench V1 Clean RerunsScience68.468.1−0.3
VoxelBench Text RatingGames23002297−0.2
PG-LLM-50Science0.40.4−0.2
OSWorld-VerifiedAgents83.483.2−0.2
LMArena Agent Arena Tool HallucinationAgents0.91.00.0
LegalBenchLegal87.087.00.0
ARC-AGI-1 Semi-PrivateGeneral QA97.597.50.0
BabaIsBench Stage 0: The IntroGames100.0100.00.0
FrontierMath ErdősMath0.00.00.0
Harvey LAB: In-House Contracting (held-out)Legal2.02.00.0
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 - 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 - 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
MMLU-Bio (SecureBio)Science91.191.10.0
Senior SWE-BenchCode34.734.70.0

Claude Opus 5vsGPT-5.6 Solnormalised score marginsData Nov 15, 2026Data: Vals AI, benchmarks.bio, Artificial Analysis and 100 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…