Atlas

Compare Models

Claude Opus 5.5 vs Gemini 4 Argon

Claude Opus 5.5 leads by 1.8 capability points (≈95% interval −0.8 to +4.4): statistically tied · P(higher) ≈ 92%

Wins 189 of 302 shared benchmarks (97 lost, 16 tied)

Capability vs. relative cost
Claude Opus 5.5Gemini 4 ArgonClosed 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 5.5vsGemini 4 Argoncapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins189 – 97 · 302 shared, 16 tied

Comparing matching setups only (46 benchmarks excluded) · suites counted once
← Claude Opus 5.5Gemini 4 Argon →

Claude Opus 5.5vsGemini 4 Argonwins 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 Gemini 4 Argon minus Claude Opus 5.5.

VCT-Refusal (CAIS)Safety90.410.7−79.7
HPCT Refusal (CAIS)Safety100.029.0−71.0
AA-PersonalityBench Transparency indexChat & Writing90.521.9−69.2
TxBench Antibody Discovery V1 (100 evaluations)Science8.363.7+55.3
LisanBench Path LengthGames240297560−51.2
TxBench Oligo Discovery V1 (113 evaluations)Science1.551.0+49.6
AA-PersonalityBench Imagination indexChat & Writing94.346.7−48.1
AA-PersonalityBench Slop indexChat & Writing20.067.6+48.1
BioSecBench-Refusal V1 Clean RerunsSafety3.848.7+44.9
SAFE-Teen 106 worst branch: Relationships (no prompt)Safety25.365.7−40.4
Harvey's Legal Agent Benchmark - Capital MarketsLegal0.040.0+40.0
Harvey's Legal Agent Benchmark - Emerging Companies Venture CapitalLegal0.040.0+40.0
Harvey's Legal Agent Benchmark - Funds Asset ManagementLegal0.040.0+40.0
Harvey's Legal Agent Benchmark - Trusts Estates Private ClientLegal0.040.0+40.0
Time Horizon Index: KSPAgents91.352.5−38.8
TxBench Preclinical Pharmacology V1 Clean RerunsScience26.364.7+38.3
BrokenArXiv 10/2026 (Native Harnesses)Math77.839.8−38.0
LLM Creative Story WritingChat & Writing3.70.8−33.0
HieroglyphBenchMultimodal37.469.7+32.3
HoneyBench v0.2 (hack rate)Safety31.663.6−32.0
BioSecBench-Function V1 Clean RerunsScience73.644.4−29.2
WitnessBench v2.0Games46.517.9−28.6
MASKSafety89.461.0−28.4
Humanity's Last Exam RMS Calibration Error (CAIS)General QA13.841.5−27.7
BrokenArXiv 09/2026 (Native Harnesses)Math76.749.7−27.0
EQ-Bench 4Chat & Writing13401236−27.0
SAFE-Teen 106 worst branch: Health Advice (no prompt)Safety44.370.8−26.5
SAFE-Teen 106 worst branch: Urgent Safety (no prompt)Safety26.552.6−26.1
Terminal-Bench 2.1Code83.557.7−25.8
SAFE-Teen Health Advice (no prompt)Safety32.856.2−23.4
BrokenArXiv 08/2026 (Native Harnesses)Math77.454.2−23.2
AA-PersonalityBench Organization indexChat & Writing79.156.2−23.1
sudo L7Code41.718.9−22.8
Machiavelli Deception (CAIS)Safety77.585.0−22.7
LisanBench Difficulty-WeightedGames59633849−22.6
SAFE-Teen 106: Health Advice (no prompt)Safety35.057.4−22.4
Code2Video-AA EloCode15361259−22.4
Agent Initiative BenchmarkAgents7.85.6−22.4
ScBench-Long V1 Clean RerunsScience6.828.4+21.6
CheatBenchSafety7.228.8−21.6
FrontierCode 2.0 - Pass RateCode51.830.6−21.2
SAFE-Teen 106: Urgent Safety (no prompt)Safety12.733.1−20.4
SAFE-Teen Urgent Safety (no prompt)Safety13.133.3−20.2
VibeVideoBench (overall quality, no video = 0)Code4.13.1−20.1
MirrorCode (ML, +Private, 2L)Code78.958.9−20.0
Harvey's Legal Agent Benchmark - Antitrust CompetitionLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Corporate GovernanceLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Corporate M&ALegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Data Privacy CybersecurityLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - ImmigrationLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Intellectual PropertyLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - International Trade SanctionsLegal0.020.0+20.0
Visa Agentic Commerce BenchmarkAgents62.282.2+20.0
TextQuests Game ProgressGames57.377.3+20.0
FrontierCode 2.0 - ScoreCode47.327.6−19.7
SAFE-Teen 106 worst branch: Urgent Safety (under-18 prompt)Safety10.530.2−19.7
FrontierML ($200 wallet)Agents58.839.5−19.3
FrontierPhysics v0.1Science27.48.3−19.1
APEX-SWECode67.649.2−18.4
Harvey HLAB — Criteria Pass RateLegal75.894.0+18.3
LMArena Agent Arena Bash RecoveryAgents15.0-21.4−18.2
HLAB-Vals-1.1Legal9.227.1+17.9
Legacy-Bench (Fireworks, avg@3)Code62.044.3−17.7
Real-SWE 1.1Code56.839.2−17.6
OpenWorldEngineer (Mean@8)Code41.624.1−17.5
LMArena Code Arena WebDevCode18141682−17.3
RWE-benchAgents29.012.0−17.0
FrontierMath Tier 4 (v2) - PrivateMath95.078.0−17.0
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal10.827.4+16.6
LMArena Agent Arena Net Improvement (Steering Burden)Agents19.510.7−16.6
DAYJOB: AccountingFinance & Accounting34.418.0−16.4
SAFE-Teen 106: Relationships (no prompt)Safety12.228.5−16.3
SAFE-Teen 106 worst branch: Health Advice (under-18 prompt)Safety27.743.6−15.9
HANDBOOK.md Agents v1.1Professional Work52.336.6−15.7
SAFE-Teen Relationships (no prompt)Safety13.028.6−15.6
DAYJOB: LawLegal25.19.9−15.2
Creative Writing v3 EloChat & Writing20511758−14.8
SnorkelLegalLegal31.446.0+14.6
SAFE-Teen Health Advice (under-18 prompt)Safety15.129.6−14.5
BioSecBench-Structure V1 (58 evaluations)Science29.314.9−14.4
Agent Red Teaming k=100 (CAIS)Safety24.410.2+14.2
SWE-bench Multimodal (Vals v1, 480 Tasks)Code70.256.0−14.2
DAYJOB: Data ScienceProfessional Work37.323.3−14.0
DAYJOB: HealthcareHealth & Medicine24.710.8−13.9
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine27.813.9−13.9
Robo-benchAgents14.60.8−13.8
WARC-Bench (192-task test split)Agents90.376.6−13.7
SAFE-Teen 106: Health Advice (under-18 prompt)Safety17.330.8−13.5
DeepFinanceFinance & Accounting54.841.4−13.4
DataCenterBench Tier 1Agents21.735.0+13.3
ChessBench CoherenceGames80.192.6+12.5
SAFE-Teen 106 worst branch: Relationships (under-18 prompt)Safety2.414.0−11.6
Senior SWE-BenchCode39.327.7−11.6
BioTIER RefusalSafety98.787.2−11.5
BrowseComp Extended (Mercor held-out)Search60.572.0+11.5
ERQAMultimodal75.186.5+11.4
Dual-Arm Sim (task completion)Agents0.011.3+11.3
ACCoRDGeneral QA88.477.3−11.1
WeirdML v3Science31.220.3−10.9
Furniture Assembly BenchmarkMultimodal83.372.5−10.8
AA-Briefcase v1.1 EloAgents16821485−10.8
Humanity's Sixth SenseMultimodal44.655.4+10.8
Ramp Accounting BenchFinance & Accounting50.561.1+10.6
LMArena Agent Arena Bash ReliabilityAgents15.1-5.4−10.2
EpiBench V1 Clean RerunsScience39.028.9−10.1
CloudOpsBench v1.0Agents47.857.5+9.8
Design Arena Overall Code CategoriesCode13841326−9.6
PCT Political Inconsistency (CAIS)Safety29.138.7−9.6
o11y-bench 1.1 Pass^3Agents93.884.4−9.4
CUA-benchAgents14.04.8−9.2
SnorkelManufacturingProfessional Work10.419.5+9.1
SAFE-Teen Urgent Safety (under-18 prompt)Safety3.312.3−9.0
MedCodeHealth & Medicine49.858.8+9.0
SAFE-Teen 106: Urgent Safety (under-18 prompt)Safety3.112.1−8.9
DualEntry Accounting AI BenchmarkFinance & Accounting82.291.1+8.9
FrontierMemory v1.1Agents73.364.4−8.9
SpatialBench-Long V1 Clean RerunsScience20.228.6+8.3
LMArena Agent Arena Praise vs ComplaintAgents40.324.1−8.1
AA-3D v1.0 EloCode12021147−8.1
ChessBench EloGames13491485+8.1
Concept Synth INDUCTION Challenge100 Round 1General QA65.057.0−8.0
Chess Puzzles (task version 2.0)Games67.059.0−8.0
GDP.xlsxProfessional Work30.338.3+8.0
HealthBench (length-adjusted, Artificial Analysis)Health & Medicine57.049.1−7.9
BioSecBench-RespiratoryVirus-Surveillance V1 (110 evaluations)Science25.417.6−7.9
SAGEMath45.853.6+7.8
AA-PersonalityBench Informativeness indexChat & Writing90.582.9−7.7
ArXivLean 09/2026Math47.239.6−7.6
FrontierSWE V2 Worst@5Code52.044.5−7.5
Conceptual Reasoning IndexIndexes77.972.4−7.5
AA-AnalystAgent (pass^5)Professional Work56.348.8−7.5
Terminal-Bench 4.0Code64.857.4−7.4
DAXBench v2Code95.788.3−7.4
PLE (Physical-Engineering's Last Exam)Professional Work50.643.2−7.4
SWE Atlas v1.1 - OverallCode77.470.1−7.3
FrontierSWE V2 Best@5Code71.564.3−7.2
Physical-Engineering's Last Exam (hidden test)Professional Work48.341.1−7.2
HealthBench ProfessionalHealth & Medicine71.464.2−7.2
BuseyBench SVGMultimodal6.87.6+7.1
CrossSearch (pass^3)Search57.064.0+7.0
Mystery Game PuzzlesGames71.064.0−7.0
VIALSMultimodal40.247.2+7.0
DistressBenchSafety75.668.6−6.9
EnigmaEvalGeneral QA57.864.6+6.8
Snorkel WorkplaceAgentsProfessional Work14.421.2+6.8
Vals RSI Index v1.1 (four tasks)Agents37.330.6−6.8
EBR-bench v4 (card ban, single agent)Games71.464.8−6.7
WCB (SecureBio)Science54.561.1+6.6
TextQuests Average Cumulative HarmSafety24.226.1−6.5
AA-Robotics v0.1Agents48.855.3+6.5
BigFinanceBench private holdout (final-answer accuracy)Finance & Accounting48.254.7+6.5
RiemannBenchMath69.676.0+6.4
IntPhys 2Multimodal68.374.6+6.4
Concept Synth INDUCTION Challenge64 Round 1General QA71.965.6−6.2
APEX-Agents 1.1Professional Work81.387.4+6.1
ClockBench AccuracyMultimodal77.283.3+6.1
GBENCH One-shot Coding GScoreGames73.967.9−6.0
APEX-Accounting 1.1Finance & Accounting65.571.5+6.0
Eval Connections One-Shot with TrapsGames86.092.0+6.0
DAYJOB: TaxFinance & Accounting9.23.2−6.0
SpeciEval (Pooled English Runs)Safety87.281.3−6.0
SnorkelRevOpsProfessional Work12.518.4+5.9
HackerRank-BenchCode51.245.4−5.9
AA-PersonalityBench Conviction indexChat & Writing69.563.8−5.8
AA-PersonalityBench Warmth indexChat & Writing84.879.1−5.8
AA-PersonalityBench Expressiveness indexChat & Writing90.584.8−5.8
ArXivMath 09/2026 (Native Harnesses)Math82.187.7+5.7
GameBench 2Games94.889.2−5.6
MBCT (SecureBio)Science68.763.2−5.5
PubMedQA (SecureBio)Health & Medicine87.982.4−5.5
MCP Atlas April 2026 (all 1,000 tasks)Agents85.490.7+5.3
Access to Justice BenchmarkLegal73.868.5−5.3
FrontierMath Tiers 1-3 (v2) - PrivateMath91.286.0−5.3
Terminal-Bench-Science 0.1Science43.348.6+5.2
DAYJOB: FinanceFinance & Accounting23.918.8−5.1
SciCode 1.0.1 (Artificial Analysis)Code66.961.8−5.1
AA-LCR v1.1Search84.779.7−5.0
BioResearchBenchScience18.013.0−5.0
CursorBench 4.0Code57.852.9−4.9
TreasuryQA BenchProfessional Work64.860.0−4.8
ScarfBench Smoke ScoreCode70.575.2+4.7
Snorkel Digital Work IndexIndexes23.628.3+4.7
FrontierCode 1.1 Extended - ScoreCode65.360.6−4.7
GDPxProfessional Work12.07.3−4.7
FrontierCode 1.1 Extended - Pass RateCode70.866.1−4.7
ChessBenchGames65.370.0+4.7
Decision-Bench v1.1Professional Work59.154.5−4.6
WMDP-Bio (SecureBio)Science96.191.5−4.6
SpatialViz-BenchMultimodal89.794.2+4.6
CritPt (avg@5)Science31.727.1−4.6
LMCAGeneral QA68.263.7−4.5
MedXpertQA MM Extended (Mercor held-out)Health & Medicine71.266.7−4.5
STEMBench120 v1.0.0 (closed-book)Science2.62.4−4.4
LegalReasonBenchLegal46.550.8+4.3
HealthBenchHealth & Medicine64.360.0−4.2
Remote Labor Index (RLI)Professional Work21.317.1−4.2
RSI-Exam 0.1 (Full 88)Agents0.50.5−4.1
Immigration BenchLegal72.168.1−4.1
MedScribeHealth & Medicine91.487.4−4.0
TaxCalcBench TY25 (web search)Finance & Accounting70.074.0+4.0
OncologyBench Strict (25 tasks, October 9 rubric)Health & Medicine29.333.3+4.0
BioSecBench-Surveillance V1 Clean RerunsScience55.859.8+4.0
BioSimBench 1.0Science14.518.4+3.9
SimpleQA Verified (Epoch anti-abstention)General QA72.276.0+3.8
ScBench V1 Clean RerunsScience55.459.1+3.8
SpatialBench V1 Clean RerunsScience72.869.0−3.8
HardGeoBench Country AccuracyMultimodal67.563.8−3.8
Arena Alignment IndexSafety83.279.4−3.7
LMArena Agent Arena Confirmed SuccessAgents20.713.2−3.7
DeepSWE v1.1Code74.277.9+3.7
Longform Creative WritingChat & Writing84.680.9−3.7
MineBench — Bradley-Terry RatingGames21342195+3.7
MindCube TinyMultimodal93.196.7+3.7
SnorkelFinance 2.0Finance & Accounting18.615.1−3.5
ArXivMath 08/2026 (Native Harnesses)Math86.082.5−3.5
ComplexConstraintsChat & Writing49.953.3+3.4
GDPval-AA v2.1Professional Work17091634−3.3
Agents' Last Exam v1.1 (ALE v1.1)Agents43.440.2−3.3
Blueprint-Bench 2Multimodal0.50.5+3.2
GDP.pdf (Artificial Analysis, All-pass)Professional Work28.825.6−3.2
VCT v2 (SecureBio)Science60.557.3−3.2
ArXivMath 10/2026 (Native Harnesses)Math79.883.0+3.2
CritPt v2 (agentic, avg@5)Science76.072.9−3.1
τ³-Banking (Mercor run, bm25_grep retrieval)Agents55.058.1+3.1
Financial Coaching BenchFinance & Accounting76.273.1−3.0
BullshitBench v2 LeaderboardSafety63.066.0+3.0
HiL-Bench (Human-in-Loop Benchmark)Agents59.062.0+3.0
MetagenomicsBench V1Science39.342.3+3.0
VCT (SecureBio)Science61.458.4−3.0
WellnessBenchSafety53.050.0−3.0
LMArena Text Arena EloChat & Writing15061522+2.8
LMArena Agent Arena Steering BurdenAgents18.613.2−2.7
ChessBench AccuracyGames79.381.8+2.7
SAFE-Teen 106: Relationships (under-18 prompt)Safety0.43.0−2.5
Terminal-Bench 4.0 (Artificial Analysis)Code52.555.1+2.5
HardGeoBench <=100miMultimodal61.358.8−2.5
HardGeoBench <=10miMultimodal42.545.0+2.5
SWE-Marathon v1.1Code46.343.8−2.5
SAFE-Teen Relationships (under-18 prompt)Safety0.53.0−2.5
LiveBench 2.0Agents86.584.1−2.4
GDPval (Mercor 213-task rubric score)Professional Work88.385.9−2.4
AML BenchFinance & Accounting66.464.1−2.3
τ³-Banking (Fireworks, banking_knowledge)Agents56.058.2+2.2
HardGeoBench Avg DistanceMultimodal669741−2.2
Vectara Hallucination LeaderboardSafety9.912.1−2.2
micro1 PharmaBenchHealth & Medicine10.88.6−2.2
Creative Writing v3 Rubric ScoreChat & Writing16.015.6−2.1
CWE-bench v2 Pass@1Cybersecurity67.165.0−2.1
ARC-AGI-2General QA93.391.3−2.1
AA-Omniscience IndexIndexes46.442.4−2.0
Bio Capabilities Index (SecureBio)Indexes171174+2.0
Toloka Arena (v3)Agents75.073.0−2.0
LAB-Bench ProtocolQA (SecureBio)Science94.095.9+1.9
SAGE v1.1Math91.389.5−1.8
Professional Reasoning Benchmark - LegalLegal47.649.4+1.8
Humanity's Last Exam DiamondGeneral QA55.053.2−1.8
MMMU ProMultimodal86.688.3+1.7
Epoch Capabilities Index (ECI)Indexes169167−1.7
BenchBench-CloningScience20.518.8−1.7
GPQA-Bio (SecureBio)Science90.989.3−1.6
AutoResearchExam (hidden-test AUARC)Agents0.60.6−1.5
FrontierMath ErdősMath2.91.5−1.5
LLM Chess EloGames15751609+1.4
Snorkel Agentic Coding 2.0Code40.441.8+1.4
Westworld Finance Diligence BenchFinance & Accounting59.460.7+1.4
Agents' Last Exam (ALE)Agents38.239.5+1.3
DTBenchGeneral QA98.997.6−1.3
Excel Modeling Benchmark v1.1Finance & Accounting75.874.8−1.0
CWE-bench v1 Programmatic Pass@1Cybersecurity67.068.0+1.0
ProofBench v1.1Math100.099.0−1.0
Bug Hunt BenchCode39.738.7−1.0
Excel Modeling BenchmarkFinance & Accounting75.975.2−0.7
Political Science LLM BenchmarkProfessional Work70.269.6−0.6
GDPval Extended (Mercor held-out rubric score)Professional Work75.876.4+0.6
MMLU-Bio (SecureBio)Science92.792.1−0.6
SnorkelUnderwrite 2.0Finance & Accounting22.723.3+0.6
MissionReportProfessional Work63.564.1+0.6
AA-LCR Extended (Mercor held-out)Search74.274.7+0.5
Humanity's Last Exam (Text-Only)General QA57.557.1−0.4
Snorkel SWE-bench CLICode16.616.3−0.3
MacroscopeBenchCode73.172.8−0.3
VariantBench V1 Clean RerunsScience50.650.3−0.3
Prophet Arena Arena Score (Fixed Context)Forecasting85.585.3−0.2
Prophet Arena Arena Score (Agentic, OpenCode)Forecasting86.286.0−0.1
LMArena Agent Arena Tool HallucinationAgents0.70.5−0.1
ALE-BenchCode274827470.0
AgenticRL-Bench (Verifiable)Agents14.014.00.0
ARC-AGI-1 Semi-PrivateGeneral QA98.598.50.0
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 - 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 - Healthcare Life SciencesLegal20.020.00.0
Harvey's Legal Agent Benchmark - InsuranceLegal0.00.00.0
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 - White Collar Defense InvestigationsLegal0.00.00.0
OTIS Mock AIME 2024-2025Math100.0100.00.0

Claude Opus 5.5vsGemini 4 Argonnormalised 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…