Atlas

Compare Models

Claude Opus 5 vs GPT-5.6 Terra

Claude Opus 5 leads by 3.4 capability points (≈95% interval +1.2 to +5.6): a clear lead over GPT-5.6 Terra · P(higher) ≈ 100%

Wins 191 of 258 shared benchmarks (44 lost, 23 tied)

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

Benchmark wins191 – 44 · 258 shared, 23 tied

Comparing matching setups only (108 benchmarks excluded) · suites counted once
← Claude Opus 5GPT-5.6 Terra →

Claude Opus 5vsGPT-5.6 Terrawins by categoryData Nov 15, 2026Data: Vals AI, Artificial Analysis, Mercor and 67 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is GPT-5.6 Terra minus Claude Opus 5.

DuetBench–Diagnosis EloHealth & Medicine1035952−99.0
HPCT Refusal (CAIS)Safety100.025.0−75.0
DecepEval Long-horizon (Neutral)Safety58.35.5+52.8
Machiavelli Deception (CAIS)Safety96.580.4+48.7
YC-BenchGames1911592636144−44.5
LisanBench Difficulty-WeightedGames50681197−41.5
LisanBench Path LengthGames154283303−37.7
DecepEval Coding (Induced)Safety64.327.3+37.0
Document ArenaProfessional Work15091468−36.7
BioTIER RefusalSafety100.065.6−34.4
APEX-SWECode63.731.1−32.6
GameBench 2Games84.853.5−31.3
ARC-AGI-3 (Semi-Private)General QA30.20.5−29.7
LMArena Agent Arena Net ImprovementAgents9.50.6−29.4
Medical Long Context Reasoning (MLCR-AA)Health & Medicine59.431.7−27.8
Terminal-Bench 4.0Code53.926.3−27.7
SM BenchSafety85.859.0−26.8
IOICode91.765.3−26.4
ProofBench v1.1Math99.074.0−25.0
Mystery Game PuzzlesGames59.035.0−24.0
SWE Atlas Codebase QnA Extended (Mercor held-out)Code40.016.0−24.0
HardGeoBench Avg DistanceMultimodal12181991−23.8
Vending-Bench 2Games111827343−23.2
MindCube TinyMultimodal84.362.6−21.7
ITSMBenchAgents50.629.0−21.6
EnigmaEvalGeneral QA43.923.1−20.8
Humanity's Last ExamGeneral QA56.435.9−20.5
TxBench Antibody Discovery V0Science51.431.0−20.4
OSWorld 2.0 500-step Partial ScoreAgents70.650.2−20.4
MedCodeHealth & Medicine63.643.4−20.2
Harvey's Legal Agent Benchmark - Corporate M&ALegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Energy Natural ResourcesLegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Healthcare Life SciencesLegal20.00.0−20.0
BullshitBench v2 LeaderboardSafety73.054.0−19.0
VCT-Refusal (CAIS)Safety29.410.7−18.7
BrowseComp Extended (Mercor held-out)Search44.025.3−18.7
AA-Omniscience IndexIndexes37.10.1−18.5
AA-Briefcase EloAgents16451330−18.3
ScBench-LongScience41.323.8−17.5
Terminal-Bench Science 0.1 (Vals)Science27.110.0−17.1
LMArena Agent Arena Net Improvement (Steering Burden)Agents7.2-1.8−17.1
Bio Capabilities Index (SecureBio)Indexes178155−17.0
SimpleQA VerifiedGeneral QA59.943.2−16.7
ALE-BenchCode26222154−16.6
HackerRank-BenchCode42.626.1−16.5
Humanity's Last Exam DiamondGeneral QA44.828.8−16.0
Extended NYT Connections 940 (Quadratic v1)Games94.378.4−15.9
GDPval Extended (Mercor held-out rubric score)Professional Work62.947.2−15.7
Hedge-BenchFinance & Accounting29.013.8−15.2
Terminal-Bench-Science 0.1Science28.613.8−14.8
Public Benefits Bench v1.1Professional Work76.962.4−14.5
NoCoT-Bench NCRI 15.2General QA11589.0−14.4
APEX-AccountingFinance & Accounting54.540.3−14.2
Creative Writing v3 EloChat & Writing21351856−14.1
GeneBench-Pro Public Release SubsetScience22.08.0−14.0
Legal Research BenchLegal55.341.3−13.9
Vibe Code Bench v1.1Code88.474.6−13.8
Design Arena Overall Code CategoriesCode13211238−13.8
GDPval-AA v2 (normalized)Professional Work67.954.1−13.8
Odysseys2Agents58.344.6−13.7
A-Fantasia (Valid Attempts with Retries)Games29.743.3−13.7
Terminal-Bench 4.0 (Artificial Analysis)Code49.035.4−13.6
WeirdML (v2)Science91.878.3−13.5
DeepsecBenchCode32.619.2−13.4
DeepsecBench — Revalidated V1Code32.419.1−13.3
DeepsecBench — Revalidated V1 (233 issues)Code32.319.1−13.3
MMMU-Pro Extended (Mercor held-out)Multimodal34.347.4+13.1
MedXpertQA MM Extended (Mercor held-out)Health & Medicine64.151.0−13.1
DeepSWE v1.1Code73.260.2−13.0
TextQuests Game ProgressGames53.340.5−12.8
Conceptual Reasoning IndexIndexes74.765.7−12.3
AutomationBench 1.0.6 (Zapier)Agents26.914.8−12.2
BabaIsBench Stage 0: The IntroGames100.088.0−12.0
CadQueryEval (Corrected Geometry)Code80.068.0−12.0
Eval Connections One-Shot with TrapsGames74.086.0+12.0
Artificial Analysis Intelligence Index v5.0.1Indexes44.232.3−11.8
GDPval-AA v2.1Professional Work17291461−11.8
BenchCAD - geometric overlapMultimodal89.978.2−11.7
Ramp SWE-Bench (78 published tasks)Code88.576.9−11.5
Toloka Arena (v1)Agents52.941.4−11.5
Tax Agent BenchFinance & Accounting76.765.2−11.5
CVDP (517-problem set)Code62.250.9−11.3
GDPval-AA v2Professional Work17081479−11.3
Agent Red Teaming k=100 (CAIS)Safety43.354.6−11.3
MRCR v2 (8-needle, Context Arena) - 128K pointwiseGeneral QA91.380.2−11.1
NoCoT-Bench NCKI (kspine v3)General QA119104−10.8
VariantBenchScience49.438.7−10.7
SpatialViz-BenchMultimodal75.865.2−10.6
HPCT (SecureBio)Science78.668.2−10.4
BALROGGames63.453.2−10.2
SpeciEval (Refusal-Aware Retries, Historical)Safety87.998.1+10.1
SpeciEval (Pooled English Runs)Safety87.998.1+10.1
HardGeoBench <=100miMultimodal46.336.3−10.0
AutomationBench (Overall)Agents25.115.2−9.9
SUPERChemScience74.764.9−9.8
HLE Extended (Mercor held-out)General QA61.070.7+9.7
Code MigrationCode57.547.8−9.7
LMCAGeneral QA64.555.0−9.5
BioSecBench-Function V0Science43.033.5−9.5
DAYJOB: FinanceFinance & Accounting11.31.8−9.5
PCT Political Inconsistency (CAIS)Safety36.746.2−9.5
DecepEval Tool-use (Induced)Safety36.045.5−9.5
DecepEval Coding (Neutral)Safety20.411.1+9.3
ACCoRDGeneral QA83.874.7−9.2
Shroombench (mushroom identification, top-1)Multimodal44.035.0−9.0
WCB (SecureBio)Science68.859.8−9.0
EnterpriseOps-Gym-AAAgents47.538.5−9.0
DeepResearch Bench II (Mercor 130-task set)Search56.147.2−8.9
TxBench Oligo Discovery V0Science42.233.3−8.9
BioSecBench-RespiratoryVirus-Surveillance V1 (110 evaluations)Science14.65.8−8.8
HardGeoBench Country AccuracyMultimodal53.845.0−8.7
LMArena Agent Arena Confirmed SuccessAgents10.2-7.2−8.7
Artificial Analysis Intelligence Index v4.3.2Indexes50.842.1−8.7
Artificial Analysis Intelligence Index v4.3Indexes50.742.3−8.4
Longform Creative WritingChat & Writing86.378.0−8.3
MedScribeHealth & Medicine91.082.9−8.1
MRCR v2 (8-needle, Context Arena) - 128K AUCGeneral QA97.589.4−8.0
ChessBench — Historical Output RepairGames32.740.7+8.0
Harvey HLAB — Criteria Pass RateLegal87.779.8−7.9
CorpFin v2Finance & Accounting73.265.3−7.9
Artificial Analysis Agentic IndexIndexes55.347.4−7.9
Terminal-Bench 2.1Code88.080.1−7.9
Humanity's Sixth SenseMultimodal30.522.7−7.8
Olam Social Poker EloGames15181511−7.6
HardGeoBench <=10miMultimodal32.525.0−7.5
LAB-Bench CloningScenarios (SecureBio)Science100.092.5−7.5
Agents' Last Exam Overall ScoreAgents55.948.5−7.4
Excel Modeling BenchmarkFinance & Accounting73.666.2−7.4
Senior SWE-BenchCode34.727.4−7.3
FrontierCode 2.0 (Artificial Analysis)Code31.223.9−7.3
ERQAMultimodal70.863.6−7.3
LiveBench 2.0Agents82.175.0−7.1
LMArena Agent Arena Praise vs ComplaintAgents12.2-1.9−7.0
ProofBenchMath78.071.0−7.0
ZeroBench (Main questions, pass@5)Multimodal26.019.0−7.0
DAYJOB: HealthcareHealth & Medicine8.41.6−6.8
Wiz Cyber Model Arena v3.9Cybersecurity71.264.4−6.8
Furniture Assembly BenchmarkMultimodal60.854.2−6.7
ARC-AGI-2General QA90.483.9−6.5
LMArena Agent Arena Bash RecoveryAgents13.60.9−6.3
BioMysteryBench - Human SolvableScience90.183.8−6.3
DistressBenchSafety70.276.4+6.2
SpatialBench-Long V0 VerifiedScience31.837.9+6.1
HieroglyphBenchMultimodal14.620.5+5.9
ChessBench CoherenceGames63.869.7+5.9
ChessBench EloGames826925+5.9
Ruby LLM Benchmarks — Program FixerCode79.874.0−5.9
SAB RankingForecasting-0.1-0.0+5.5
CursorBench 4.0Code46.641.3−5.3
APEX-Agents 1.1Professional Work77.071.8−5.2
CursorBench 3.2Code70.064.9−5.1
WordleBenchGames96.091.0−5.0
Roboflow Vision Evals (October 2026, five tasks)Multimodal70.966.0−4.9
BioSecBench-RefusalSafety29.434.3+4.9
GDPval (Mercor 213-task rubric score)Professional Work85.680.7−4.9
Vision Arena OverallMultimodal12851266−4.9
GBENCH One-shot Coding GScoreGames68.163.3−4.8
DecepEval Tool-use (Neutral)Safety9.34.5+4.8
MortgageTax (Vals all 1,258 documents)Finance & Accounting72.167.3−4.7
EpiBenchScience35.239.9+4.7
Concept Synth INDUCTION Challenge64 Round 1General QA32.837.5+4.7
DTBenchGeneral QA97.993.3−4.5
𝜏³-BankingAgents44.740.2−4.5
LMArena Agent Arena Bash ReliabilityAgents9.80.7−4.5
MMMU ProMultimodal84.079.5−4.5
Roboflow Vision Evals (September 2026 protocol)Multimodal78.373.8−4.5
ReactBench v1Code49.053.3+4.3
VCT (SecureBio)Science51.447.2−4.2
LAB-Bench ProtocolQA (SecureBio)Science81.177.0−4.1
MedXpertQA MM (200-question set)Health & Medicine79.875.7−4.1
AA-LCRSearch70.074.0+4.0
Chess Puzzles (task version 2.0)Games43.047.0+4.0
Finance Agent v2 (Vals Index subset)Finance & Accounting56.652.6−3.9
Finance Agent v2 (Vals Multimodal Index component)Finance & Accounting56.652.6−3.9
MMLU ProGeneral QA92.088.1−3.9
IntPhys 2Multimodal65.161.3−3.8
LMArena Text Arena EloChat & Writing14901468−3.8
RedlineBenchLegal44.047.8+3.8
SpatialBench V0 VerifiedScience66.770.4+3.8
Artificial Analysis Progress Index v1Indexes334324−3.5
MBCT (SecureBio)Science60.256.9−3.3
ProgramBench (Mercor 100-task subset)Code4.31.0−3.3
SWE-bench Verified Extended (Mercor held-out)Code82.078.7−3.3
LMArena Agent Arena SteerabilityAgents10.74.1−3.3
IOI (Vals v2)Code84.387.6+3.3
Political Science LLM BenchmarkProfessional Work71.468.3−3.2
LiveCodeBench (Vals AI public dataset)Code89.085.9−3.1
AutomationBench-AA 1.0.6Agents56.659.6+3.1
AA-LCR v1.1Search82.079.0−3.0
GDP.pdf (Artificial Analysis, All-pass)Professional Work21.624.6+3.0
ZeroBench Main Mean pass@1 (five samples)Multimodal17.214.2−3.0
Epoch Capabilities Index (ECI)Indexes164161−2.9
SAB ResearchFinance & Accounting96.293.4−2.8
GPQA-Bio (SecureBio)Science82.785.3+2.6
GPQA DiamondGeneral QA92.489.9−2.5
LMArena Agent Arena Steering BurdenAgents4.4-0.5−2.5
FrontierMath Tier 4 (v2) - PrivateMath73.270.7−2.4
SAGEMath49.447.0−2.4
AdvancedIFChat & Writing80.578.1−2.4
BuseyBench SVGMultimodal6.36.1−2.2
ObviousBenchGeneral QA100.097.9−2.1
LegalBenchLegal87.085.1−1.9
Creative Writing v3 Rubric ScoreChat & Writing16.315.9−1.8
SciCode (Artificial Analysis avg@3)Science55.753.9−1.7
CritPt v2 (agentic, avg@5)Science71.469.7−1.7
Terminal-Bench 2.1 Extended (Mercor held-out)Code38.036.4−1.6
RiemannBenchMath68.066.4−1.6
SWE-bench VerifiedCode97.095.4−1.6
ScBench-Long V0 VerifiedScience36.434.9−1.5
SciCode (main problems)Science17.315.8−1.5
DecepEval Long-horizon (Induced)Safety83.384.7−1.4
TextQuests Average Cumulative HarmSafety22.622.2+1.4
Artificial Analysis Coding IndexIndexes78.076.7−1.3
CharXiv Extended (Mercor held-out)Multimodal79.881.1+1.3
BrowseComp (Mercor 130-question subset)Search84.685.8+1.2
CRM-Agent-BenchAgents82.781.6−1.1
TaxEval v2Finance & Accounting75.176.2+1.0
ARC-AGI-1 Semi-PrivateGeneral QA97.596.5−1.0
Concept Synth INDUCTION Challenge100 Round 1General QA24.025.0+1.0
ZeroBench Main Questions pass^5Multimodal11.010.0−1.0
ChessBench AccuracyGames71.071.9+1.0
APEX-1 (Mercor, mean score)Professional Work68.669.5+0.9
CritPt (avg@5)Science29.130.0+0.9
OTIS Mock AIME 2024-2025Math98.999.7+0.8
DeepsecBench-AA (Artificial Analysis)Cybersecurity25.925.1−0.8
AA-LCR Extended (Mercor held-out)Search73.974.4+0.5
CharXiv (descriptive and reasoning combined)Multimodal94.694.1−0.5
Blueprint-Bench 2Multimodal0.30.3+0.4
MMLU-Bio (SecureBio)Science91.190.7−0.4
FrontierMath Tiers 1-3 (v2) - PrivateMath85.686.0+0.4
LLM Chess EloGames12851293+0.3
LAB-Bench LitQA2 (SecureBio)Science51.451.1−0.3
SWE-sweepCode1.31.5+0.3
MASKSafety93.493.3−0.1
LMArena Agent Arena Tool HallucinationAgents0.91.00.0
AIME 2025Math100.0100.00.0
BioMysteryBench - Human DifficultScience49.449.40.0
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 - 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 SecuritizationLegal20.020.00.0
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

Claude Opus 5vsGPT-5.6 Terranormalised score marginsData Nov 15, 2026Data: Vals AI, Artificial Analysis, Mercor and 67 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…