Atlas

Compare Models

Claude Opus 5.5 vs GPT-6.1 Sol

Claude Opus 5.5 leads by 1.5 capability points (≈95% interval −1.2 to +4.1): statistically tied · P(higher) ≈ 86%

Wins 156 of 292 shared benchmarks (103 lost, 33 tied)

Capability vs. relative cost
Claude Opus 5.5GPT-6.1 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-6.1 Solcapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins156 – 103 · 292 shared, 33 tied

Comparing matching setups only (138 benchmarks excluded) · suites counted once
← Claude Opus 5.5GPT-6.1 Sol →

Claude Opus 5.5vsGPT-6.1 Solwins by categoryData Nov 15, 2026Data: Vals AI, Mercor, Artificial Analysis and 76 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is GPT-6.1 Sol minus Claude Opus 5.5.

KernelBench Mega Normalized (RTX PRO 6000)Code100.014.1−85.9
VCT-Refusal (CAIS)Safety90.416.1−74.3
YC-BenchGames2873725936267−67.6
AA-PersonalityBench Informativeness indexChat & Writing90.527.7−63.5
Mercor enterprise risk (implicit-risk unsafe rate)Safety22.283.3−61.1
HPCT Refusal (CAIS)Safety100.042.0−58.0
Short Form Bench v1.0Games23892949+55.5
TxBench Oligo Discovery V1 (113 evaluations)Science1.552.5+51.0
LisanBench Path LengthGames240298736−47.5
TxBench Antibody Discovery V1 (100 evaluations)Science8.349.7+41.3
Design Creativity BenchCode51.063.7+40.4
AA-PersonalityBench Expressiveness indexChat & Writing90.550.5−40.4
Olam Social Poker EloGames15061543+40.2
Clipping Bench v1.0Games20185707+39.7
CheatBenchSafety7.242.8−35.6
SAFE-Teen 106 worst branch: Health Advice (no prompt)Safety44.38.8+35.5
BioSecBench-Refusal V1 Clean RerunsSafety3.837.6+33.8
TxBench Preclinical Pharmacology V1 Clean RerunsScience26.359.7+33.3
KernelBench CUDA Normalized (RTX PRO 6000)Code76.343.2−33.0
Medical Long Context Reasoning (MLCR-AA)Health & Medicine66.733.9−32.8
AA-PersonalityBench Imagination indexChat & Writing94.361.9−32.7
Clocktower BenchGames15471764+31.1
ActiveVisionMultimodal41.271.8+30.6
SAFE-Teen 106: Health Advice (no prompt)Safety35.04.7+30.3
Humanity's Last Exam RMS Calibration Error (CAIS)General QA13.842.5−28.6
SAFE-Teen Health Advice (no prompt)Safety32.84.2+28.6
HoneyBench v0.2 (hack rate)Safety31.659.5−27.8
BioSecBench-Function V1 Clean RerunsScience73.646.0−27.6
SWE-bench Multimodal (Vals v1, 480 Tasks)Code70.243.5−26.7
Concept Synth INDUCTION Challenge100 Round 1General QA65.090.0+25.0
EnterpriseBench: CoreCraftProfessional Work74.450.3−24.1
HANDBOOK.md Agents v1.1Professional Work52.328.6−23.7
Concept Synth INDUCTION Challenge64 Round 1General QA71.995.3+23.4
FrontierMemory v1.1Agents73.351.1−22.2
LisanBench Difficulty-WeightedGames59634004−21.0
SWE-bench Verified Extended (Mercor held-out)Code66.045.3−20.7
APEX-SWECode67.646.9−20.7
SAFE-Teen 106 worst branch: Health Advice (under-18 prompt)Safety27.77.7+20.0
Harvey's Legal Agent Benchmark - Corporate M&ALegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Healthcare Life SciencesLegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Structured Finance SecuritizationLegal0.020.0+20.0
ScBench-Long V1 Clean RerunsScience6.826.7+19.9
SpatialBench-Long V1 Clean RerunsScience20.239.9+19.6
AA-PersonalityBench Warmth indexChat & Writing84.865.7−19.2
DAYJOB: Data ScienceProfessional Work37.318.7−18.6
SpireClimbGames5.823.8+18.0
HardGeoBench Avg DistanceMultimodal6691239−17.6
EQ-Bench 4Chat & Writing13401272−17.6
HardGeoBench <=100miMultimodal61.343.8−17.5
TextQuests Average Cumulative HarmSafety24.229.3−17.4
EBR-bench v4 (card ban, single agent)Games71.454.3−17.1
DAYJOB: HealthcareHealth & Medicine24.77.6−17.1
DataCenterBench Tier 1Agents21.738.3+16.7
ViBench 2.0Code48.564.7+16.2
AA-PersonalityBench Conviction indexChat & Writing69.554.3−15.4
HardGeoBench Country AccuracyMultimodal67.552.5−15.0
TextQuests Game ProgressGames57.371.4+14.1
SAFE-Teen 106: Health Advice (under-18 prompt)Safety17.33.3+14.0
ProgramBench (Mercor 100-task subset)Code20.06.3−13.7
Search ArenaSearch12131230+13.7
Code2Video-AA EloCode15361369−13.5
AA-Briefcase v1.1 EloAgents18011555−13.5
ArXivMath 05/2026Math81.795.0+13.3
LMArena Agent Arena Net Improvement (Steering Burden)Agents19.512.5−13.3
SAFE-Teen Health Advice (under-18 prompt)Safety15.12.0+13.1
BioSecBench-RespiratoryVirus-Surveillance V1 (110 evaluations)Science25.412.4−13.0
WellnessBenchSafety53.066.0+13.0
DeepsecBench — Revalidated V1 (233 issues)Code26.639.5+12.9
SM BenchSafety81.869.0−12.8
SciCode (main problems)Science27.314.6−12.7
ChessBench EloGames13491562+12.7
Harvey HLAB — Criteria Pass RateLegal75.887.8+12.0
τ³-Banking (Mercor run, bm25_grep retrieval)Agents55.043.0−12.0
SnorkelFinance 2.0Finance & Accounting18.66.6−12.0
SWE Atlas Codebase QnA Extended (Mercor held-out)Code43.331.3−12.0
AA-3D v1.0 EloCode12021121−11.8
sudo L7 (October 2026 rescoring)Code45.633.9−11.7
SWE Atlas v1.1 - OverallCode77.465.9−11.5
Terminal-Bench 4.0Code64.853.4−11.4
Terminal-Bench-Science 0.1Science61.950.5−11.4
GDPval-AA v2.1Professional Work18661607−11.4
τ³-Banking (Fireworks, banking_knowledge)Agents56.044.7−11.3
SAFE-Teen 106 worst branch: Urgent Safety (no prompt)Safety26.515.4+11.2
SciCode 1.0.1 (Artificial Analysis)Code66.955.8−11.1
ArXivMath 06/2026Math83.394.4+11.1
BuseyBench SVGMultimodal6.88.0+11.0
SWE-bench VerifiedCode98.287.3−10.9
SnorkelLegalLegal31.420.7−10.7
EpiBench V1 Clean RerunsScience39.028.3−10.7
FrontierCUA pass@1Agents36.046.7+10.7
SWE-bench MultilingualCode95.584.9−10.6
SWE-TogetherCode68.858.3−10.5
MMMU-Pro Extended (Mercor held-out)Multimodal51.061.3+10.3
SpeciEval (Refusal-Aware Retries, Historical)Safety87.297.4+10.1
SpeciEval (Pooled English Runs)Safety87.297.4+10.1
Decision-Bench v1.1Professional Work59.149.1−10.1
FACTS Benchmark SuiteGeneral QA65.375.3+10.0
SAFE-Teen Urgent Safety (no prompt)Safety13.13.5+9.7
SAFE-Teen 106: Urgent Safety (no prompt)Safety12.73.4+9.3
SnorkelManufacturingProfessional Work10.419.6+9.2
DAYJOB: AccountingFinance & Accounting34.425.2−9.2
Mystery Game PuzzlesGames71.080.0+9.0
Toloka Arena (v3)Agents75.066.0−9.0
CVDP (517-problem set)Code64.155.1−9.0
VIALSMultimodal40.249.1+8.9
MASKSafety89.498.3+8.9
ALE-BenchCode27482990+8.6
GDP.xlsxProfessional Work30.322.0−8.3
ScBench V1 Clean RerunsScience55.463.6+8.2
PCT Political Inconsistency (CAIS)Safety29.137.3−8.2
Agent Initiative BenchmarkAgents7.88.7+8.1
APEX-Agents 1.1Professional Work81.373.2−8.1
BioSecBench-Structure V1 (58 evaluations)Science29.321.3−8.0
Access to Justice BenchmarkLegal73.865.8−7.9
Financial Coaching BenchFinance & Accounting76.268.5−7.6
Gray Swan IPI Attack Success (k=15)Safety1.08.6−7.6
ACCoRDGeneral QA88.480.9−7.6
HardGeoBench <=10miMultimodal42.535.0−7.5
ChessBench CoherenceGames80.187.5+7.4
HealthBenchHealth & Medicine64.356.9−7.3
BrowseComp Extended (Mercor held-out)Search60.567.7+7.2
Conceptual Reasoning IndexIndexes77.972.7−7.1
Extended NYT Connections 940 (Quadratic v1)Games88.595.5+7.0
Legacy-Bench (Fireworks, avg@3)Code62.055.0−7.0
BrowseComp (Mercor 130-question subset)Search88.595.4+6.9
Epiq LAER CorporateBenchProfessional Work75.482.2+6.8
Snorkel SWE-bench CLICode16.69.8−6.8
DAYJOB: LawLegal25.131.8+6.7
Finance Agent v2Finance & Accounting58.652.0−6.6
MedXpertQA MM Extended (Mercor held-out)Health & Medicine71.264.7−6.5
LMArena Agent Arena Praise vs ComplaintAgents40.327.5−6.4
MetagenomicsBench V1Science39.345.7+6.3
AA-AnalystAgent (pass^5)Professional Work56.350.0−6.2
o11y-bench 1.1 Pass^3Agents93.887.5−6.2
SAFE-Teen 106 worst branch: Relationships (no prompt)Safety25.319.1+6.2
HieroglyphBenchMultimodal37.431.4−6.0
DAYJOB: FinanceFinance & Accounting23.918.0−5.9
FrontierCode 2.0 (Artificial Analysis)Code46.440.6−5.8
AA-PersonalityBench Slop indexChat & Writing20.014.3−5.8
ComplexConstraints (November 2026 rescoring)Chat & Writing64.470.1+5.7
Agent Red Teaming k=100 (CAIS)Safety24.418.8+5.7
Terminal-Bench 4.0 (Artificial Analysis)Code59.654.0−5.6
Mercor enterprise risk (explicit-risk unsafe rate)Safety5.60.0+5.6
SimpleBenchGeneral QA88.482.9−5.5
SpatialBench V1 Clean RerunsScience72.867.3−5.5
LMCAGeneral QA68.262.8−5.5
MacroscopeBenchCode83.978.5−5.4
OncologyBench Strict (25 tasks, October 9 rubric)Health & Medicine29.324.0−5.3
LMArena Agent Arena Bash RecoveryAgents15.04.5−5.2
DAYJOB: TaxFinance & Accounting9.24.0−5.2
Excel Modeling BenchmarkFinance & Accounting75.970.8−5.2
GBENCH One-shot Coding GScoreGames73.979.0+5.1
LMArena Agent Arena Bash ReliabilityAgents15.15.0−5.1
BioSecBench-Surveillance V1 Clean RerunsScience55.850.8−5.0
BrokenArXiv 05/2026Math95.090.0−5.0
FrontierMath Tier 4 (v2) - PrivateMath95.0100.0+5.0
MedScribeHealth & Medicine91.486.5−5.0
Humanity's Last Exam (Text-Only)General QA57.552.6−4.9
Arena Alignment IndexSafety83.287.9+4.8
Excel Modeling Benchmark v1.1Finance & Accounting75.871.1−4.7
Design Arena Overall Code CategoriesCode13841356−4.6
Agents' Last Exam Overall ScoreAgents63.258.6−4.6
VariantBench V1 Clean RerunsScience50.646.0−4.5
PI OfficeQA Pro v2Professional Work85.681.1−4.5
VoxelBench Text RatingGames26132550−4.5
LMArena Agent Arena Confirmed SuccessAgents20.711.7−4.5
Westworld Finance Diligence BenchFinance & Accounting59.455.0−4.4
Immigration BenchLegal72.167.8−4.3
Humanity's Last Exam DiamondGeneral QA54.658.8+4.2
LMArena Text Arena EloChat & Writing15061482−4.2
Ramp Accounting BenchFinance & Accounting50.546.4−4.2
IntPhys 2Multimodal68.372.3+4.1
Machiavelli Deception (CAIS)Safety77.578.8−4.0
TaxCalcBench TY25Finance & Accounting42.038.0−4.0
WARC-Bench (192-task test split)Agents90.386.3−4.0
MindCube TinyMultimodal93.189.1−4.0
BigFinanceBench private holdout (final-answer accuracy)Finance & Accounting48.252.0+3.8
GDPxProfessional Work12.08.4−3.6
ChessBenchGames65.368.7+3.3
Furniture Assembly BenchmarkMultimodal83.380.0−3.3
HiL-Bench (Human-in-Loop Benchmark)Agents59.062.3+3.3
GDP.pdf (Artificial Analysis, All-pass)Professional Work28.832.0+3.2
DAXBench v2Code95.792.6−3.2
Longtail-SWECode81.578.4−3.1
HealthBench ProfessionalHealth & Medicine71.468.3−3.1
ERQAMultimodal75.178.2+3.1
SAFE-Teen Relationships (no prompt)Safety13.010.0+3.1
FrontierCode 1.1 Main (Artificial Analysis)Code52.249.1−3.1
GeneBench-Pro Public Release SubsetScience30.033.0+3.0
DeepSWE v1.1Code72.375.2+2.9
HLAB-Vals-1.1Legal9.26.3−2.9
ClockBench AccuracyMultimodal77.274.4−2.8
Hermes BenchAgents76.773.9−2.8
GameBench 2Games94.892.0−2.8
LMArena Agent Arena Steering BurdenAgents18.613.4−2.6
Ruby LLM Benchmarks — Program FixerCode77.575.0−2.5
MissionReportProfessional Work63.561.0−2.5
FrontierMath Tiers 1-3 (v2) - PrivateMath91.293.7+2.5
AA-Omniscience IndexIndexes46.441.5−2.4
Snorkel Digital Work IndexIndexes23.621.2−2.4
DTBenchGeneral QA98.996.5−2.4
MMLU ProGeneral QA93.491.0−2.4
Snorkel Agentic Coding 2.0Code40.442.8+2.4
SnorkelUnderwrite 2.0Finance & Accounting22.725.0+2.3
SAFE-Teen 106 worst branch: Urgent Safety (under-18 prompt)Safety10.512.7−2.1
SAFE-Teen 106 worst branch: Relationships (under-18 prompt)Safety2.44.5−2.1
SAFE-Teen 106: Relationships (no prompt)Safety12.210.2+2.0
CrossSearch (pass^3)Search57.059.0+2.0
BullshitBench v2 LeaderboardSafety63.065.0+2.0
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal10.88.8−2.0
Humanity's Sixth SenseMultimodal44.646.6+2.0
AA-PersonalityBench Organization indexChat & Writing79.181.0+1.9
AA-PersonalityBench Transparency indexChat & Writing90.592.4+1.9
Terminal-Bench 2.1Code83.585.4+1.9
Political Science LLM BenchmarkProfessional Work70.268.3−1.8
SUPERChemScience79.677.8−1.8
EnigmaEvalGeneral QA57.859.5+1.7
SimpleQA Verified (Epoch anti-abstention)General QA72.273.9+1.7
Terminal-Bench 2.1 Extended (Mercor held-out)Code47.545.8−1.7
Harvey LAB: In-House Contracting (held-out)Legal3.04.7+1.7
HealthBench (length-adjusted, Artificial Analysis)Health & Medicine57.058.5+1.5
MMMU ProMultimodal86.685.1−1.5
ChessBench AccuracyGames79.380.7+1.5
LLM Chess EloGames15751612+1.5
FrontierMath ErdősMath2.94.4+1.5
DistressBenchSafety75.674.2−1.4
SAFE-Teen Urgent Safety (under-18 prompt)Safety3.32.1+1.2
APEX-Accounting 1.1Finance & Accounting65.564.3−1.2
SAFE-Teen 106: Urgent Safety (under-18 prompt)Safety3.12.0+1.2
LiveBench 2.0Agents86.585.4−1.1
Hebbia Financial Services BenchmarkFinance & Accounting74.673.5−1.1
Chess Puzzles (task version 2.0)Games67.068.0+1.0
DualEntry Accounting AI BenchmarkFinance & Accounting82.281.2−1.0
ProofBench v1.1Math100.099.0−1.0
MedCodeHealth & Medicine49.848.8−1.0
CritPt v2 (agentic, avg@5)Science81.782.6+0.9
ARC-AGI-2General QA93.394.2+0.8
SAGE v1.1Math91.390.5−0.8
BioSimBench 1.0Science14.515.3+0.8
RiemannBench (November 2026 rescoring)Math80.079.2−0.8
SAGEMath45.846.5+0.7
AA-LCR v1.1Search84.784.0−0.7
Prophet Arena Arena Score (Fixed Context)Forecasting85.584.8−0.6
Snorkel WorkplaceAgentsProfessional Work14.415.0+0.6
SpatialViz-BenchMultimodal89.790.3+0.6
ForecastBench Preliminary Dataset Brier IndexForecasting62.261.7−0.5
SWE Atlas - Codebase QnACode48.948.4−0.5
OpenWorldEngineer (Mean@8)Code41.641.2−0.4
SAFE-Teen Relationships (under-18 prompt)Safety0.50.9−0.4
SAFE-Teen 106: Relationships (under-18 prompt)Safety0.40.9−0.4
CharXiv (descriptive and reasoning combined)Multimodal96.496.0−0.4
SnorkelRevOpsProfessional Work12.512.1−0.4
AdvancedIFChat & Writing82.882.4−0.4
CharXiv Extended (Mercor held-out)Multimodal84.785.1+0.4
Roboflow Vision Evals (October 2026, five tasks)Multimodal82.081.7−0.3
Professional Reasoning Benchmark - LegalLegal47.647.9+0.3
Epoch Capabilities Index (ECI)Indexes169169+0.2
LMArena Agent Arena Tool HallucinationAgents0.71.0+0.2
Prophet Arena Arena Score (Agentic, OpenCode)Forecasting86.286.0−0.1
AIME 2025Math100.0100.00.0
ARC-AGI-1 Semi-PrivateGeneral QA98.598.50.0
BrokenArXiv 06/2026Math100.0100.00.0
ClaimBenchFinance & Accounting100.0100.00.0
CritPt (avg@5)Science31.731.70.0
Dual-Arm Sim (task completion)Agents0.00.00.0
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
MedXpertQA MM (200-question set)Health & Medicine85.785.70.0
Mercor enterprise risk (benign-control completion rate)Safety100.0100.00.0
ObviousBenchGeneral QA100.0100.00.0
OTIS Mock AIME 2024-2025Math100.0100.00.0
Roboflow Vision Evals (September 2026 protocol)Multimodal85.585.50.0
Visa Agentic Commerce BenchmarkAgents62.262.20.0

Claude Opus 5.5vsGPT-6.1 Solnormalised score marginsData Nov 15, 2026Data: Vals AI, Mercor, Artificial Analysis and 76 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…