Atlas

Compare Models

Gemini 3.7 Flash vs Muse Wonder

Muse Wonder leads by 4.2 capability points (≈95% interval +2.1 to +6.3): a clear lead over Gemini 3.7 Flash · P(higher) ≈ 100%

Wins 104 of 149 shared benchmarks (31 lost, 14 tied)

Capability vs. relative cost
Gemini 3.7 FlashMuse WonderClosed 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.

Gemini 3.7 FlashvsMuse Wondercapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins31 – 104 · 149 shared, 14 tied

Comparing matching setups only (11 benchmarks excluded) · suites counted once
← Gemini 3.7 FlashMuse Wonder →

Gemini 3.7 FlashvsMuse Wonderwins by categoryData Nov 15, 2026Data: Vals AI, Arena, Center for AI Safety and 25 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Muse Wonder minus Gemini 3.7 Flash.

HPCT Refusal (CAIS)Safety13.097.0+84.0
VCT-Refusal (CAIS)Safety4.860.2+55.4
Terminal-Bench 4.0 — SoftwareCode7.457.4+50.0
Harvey's Legal Agent Benchmark - Litigation Dispute ResolutionLegal0.040.0+40.0
CyberBench v1.1 OverallCybersecurity43.880.4+36.6
Terminal-Bench 4.0 — ScienceCode11.947.6+35.7
FrontierMath Tier 4 (v2) - PrivateMath36.670.7+34.1
Terminal-Bench 4.0 — HardwareCode13.346.7+33.3
Terminal-Bench 4.0 — SecurityCode33.366.7+33.3
Terminal-Bench 4.0 — MLCode12.145.5+33.3
Terminal-Bench 4.0 — MediaCode0.033.3+33.3
DeepResearch Bench II (Mercor 130-task set)Search43.410.8−32.6
MASKSafety53.784.9+31.2
LLM Creative Story WritingChat & Writing-0.72.0+30.1
Code MigrationCode34.862.6+27.8
Furniture Assembly BenchmarkMultimodal26.754.2+27.5
AA-Briefcase v1.1 EloAgents11151610+27.1
ProofBench v1.1Math58.084.0+26.0
Terminal-Bench 4.0 — OperationsCode14.840.7+25.9
AA-LCR Extended (Mercor held-out)Search48.974.4+25.5
Terminal-Bench 4.0 (Artificial Analysis)Code13.638.4+24.7
HackerRank-BenchCode28.549.7+21.3
Harvey's Legal Agent Benchmark - Antitrust CompetitionLegal0.020.0+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 - Energy Natural ResourcesLegal40.020.0−20.0
Harvey's Legal Agent Benchmark - Funds Asset ManagementLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - InsuranceLegal20.00.0−20.0
Harvey's Legal Agent Benchmark - Intellectual PropertyLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Real EstateLegal0.020.0+20.0
Harvey's Legal Agent Benchmark - Trusts Estates Private ClientLegal0.020.0+20.0
DAYJOB: FinanceFinance & Accounting2.821.3+18.5
DAYJOB: HealthcareHealth & Medicine0.018.0+18.0
DistressBenchSafety69.085.9+16.9
ChessBench CoherenceGames83.2100.0+16.8
GDPval-AA v2.1Professional Work14041770+16.1
Legal Research BenchLegal34.650.5+15.9
Terminal-Bench Science 0.1 (Vals)Science7.122.9+15.7
NoCoT-Bench NCRI 15.2General QA12697.7−15.6
RuneBenchGames6.05.0−15.1
Arena Code WebDev — Brand & MarketingCode15531653+14.6
Creative Writing v3 EloChat & Writing17202007+14.5
BioTIER RefusalSafety84.498.7+14.3
CritPt (avg@5)Science14.328.3+14.0
Mystery Game PuzzlesGames37.051.0+14.0
Tax Agent BenchFinance & Accounting57.771.2+13.5
Artificial Analysis Intelligence Index v4.3.2Indexes39.152.4+13.3
DeepSWE v1.1Code65.578.8+13.3
ChessBench EloGames12651046−13.0
SpatialBench V1 Clean RerunsScience59.171.9+12.7
Vibe Code Bench v1.1Code70.483.1+12.7
APEX-SWECode42.054.6+12.6
Finance Agent v2 - Market AnalysisFinance & Accounting74.562.1−12.4
TextQuests Average Cumulative HarmSafety24.420.8+12.4
GDPval Extended (Mercor held-out rubric score)Professional Work57.870.0+12.2
PCT Political Inconsistency (CAIS)Safety45.133.3+11.8
Arena Code WebDev — FullstackCode16291674+11.6
LiveBench 2.0Agents67.278.1+10.9
FrontierMath Tiers 1-3 (v2) - PrivateMath71.682.1+10.5
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine12.222.2+10.0
APEX-Accounting 1.1Finance & Accounting57.267.1+9.9
Arena Code WebDev — HTMLCode15931664+9.8
Vals Index v2.1Indexes51.361.1+9.8
ChessBench AccuracyGames79.270.2−9.6
OncologyBench Strict (25 tasks, October 9 rubric)Health & Medicine14.724.0+9.3
Bug Hunt BenchCode21.030.2+9.2
Arena Code WebDev — ReactCode15841651+9.0
SpeciEval (Pooled English Runs)Safety89.097.5+8.5
Arena Code WebDev — FrontendCode15921657+8.4
SpatialViz-BenchMultimodal75.383.5+8.2
APEX-1 (Mercor, mean score)Professional Work67.075.1+8.1
NoCoT-Bench NCKI (kspine v3)General QA121109−8.1
TaxCalcBench TY25Finance & Accounting14.022.0+8.0
Arena Code WebDev — SimulationsCode16181688+7.9
Humanity's Last Exam DiamondGeneral QA31.239.0+7.8
Toloka Arena (v3)Agents58.065.7+7.7
LMArena Agent Arena Net Improvement (Steering Burden)Agents-2.21.8+7.5
Eval Connections One-Shot with TrapsGames80.073.0−7.0
GDP.pdf (Artificial Analysis, All-pass)Professional Work23.630.6+7.0
Arena Code WebDev — Reference-Based DesignCode16081660+6.9
MMMU-Pro Extended (Mercor held-out)Multimodal36.443.2+6.8
Arena Code WebDev — Consumer ProductCode15741628+6.8
LMCAGeneral QA50.456.8+6.5
Artificial Analysis Progress Index v1Indexes317336+6.5
Ramp Accounting BenchFinance & Accounting42.348.7+6.4
MineBench — Bradley-Terry RatingGames17651868+6.2
AutomationBench-AA 1.0.6Agents62.068.1+6.1
Machiavelli Deception (CAIS)Safety85.583.6+5.9
Arena Code WebDev — GamingCode16381689+5.7
EnterpriseBench: CoreCraftProfessional Work55.961.5+5.6
Finance Agent v2 - General Qualitative AnalysisFinance & Accounting65.370.8+5.5
GBENCH One-shot Coding GScoreGames52.658.1+5.5
TextQuests Game ProgressGames52.947.6−5.3
Creative Writing v3 Rubric ScoreChat & Writing15.216.2+5.0
VIALSMultimodal26.531.5+5.0
LMArena Agent Arena Steering BurdenAgents-7.32.6+4.9
LMArena Agent Arena Confirmed SuccessAgents-3.55.4+4.4
Conceptual Reasoning IndexIndexes64.667.7+4.4
Excel Modeling BenchmarkFinance & Accounting71.367.1−4.2
Finance Agent v2 - General Quantitative AnalysisFinance & Accounting81.877.6−4.2
Arena Code WebDev — Content Creation ToolsCode15741601+4.1
Harvey HLAB — Criteria Pass RateLegal88.692.7+4.1
Humanity's Sixth SenseMultimodal39.843.7+3.9
LMArena Text Arena EloChat & Writing14891511+3.8
Epoch Capabilities Index (ECI)Indexes158162+2.9
MindCube TinyMultimodal88.491.1+2.7
ERQAMultimodal77.775.3−2.5
APEX-Agents 1.1Professional Work79.477.0−2.4
Arena Code WebDev — Data & AnalyticsCode15771593+2.4
Finance Agent v2 - ComparablesFinance & Accounting50.347.9−2.4
Finance Agent v2 - Earnings AnalysisFinance & Accounting79.176.8−2.3
OTIS Mock AIME 2024-2025Math97.299.4+2.2
EnigmaEvalGeneral QA33.936.1+2.2
Finance Agent v2 - AdjustmentsFinance & Accounting54.752.7−2.0
Humanity's Last ExamGeneral QA45.647.6+2.0
IntPhys 2Multimodal68.270.2+2.0
GDPval (Mercor 213-task rubric score)Professional Work85.287.0+1.8
Finance Agent v2 - PrecedentsFinance & Accounting33.832.1−1.7
Finance Agent v2 - Financial ModelingFinance & Accounting24.222.5−1.6
DTBenchGeneral QA96.895.2−1.6
Finance Agent v2 - All PassFinance & Accounting47.545.9−1.5
MMLU ProGeneral QA90.689.6−1.0
IOI (Vals v2)Code67.866.9−0.9
ACCoRDGeneral QA77.076.2−0.8
CharXiv (descriptive and reasoning combined)Multimodal95.394.5−0.8
LMArena Agent Arena Bash ReliabilityAgents6.24.8−0.7
LMArena Agent Arena Tool HallucinationAgents0.7-0.5−0.6
LMArena Agent Arena Praise vs ComplaintAgents-2.9-1.8+0.5
BioSecBench-Refusal V1 Clean RerunsSafety54.855.3+0.5
AA-Omniscience IndexIndexes26.525.5−0.5
CharXiv Extended (Mercor held-out)Multimodal79.179.6+0.5
Finance Agent v2 - Disclosure AnalysisFinance & Accounting67.767.3−0.5
SciCode 1.0.1 (Artificial Analysis)Code59.860.3+0.5
BuseyBench SVGMultimodal6.66.6−0.4
MMMU ProMultimodal84.984.90.0
Harvey's Legal Agent Benchmark - Arbitration International Dispute ResolutionLegal0.00.00.0
Harvey's Legal Agent Benchmark - Banking FinanceLegal20.020.00.0
Harvey's Legal Agent Benchmark - Bankruptcy RestructuringLegal0.00.00.0
Harvey's Legal Agent Benchmark - Capital MarketsLegal20.020.00.0
Harvey's Legal Agent Benchmark - Corporate GovernanceLegal20.020.00.0
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 - Healthcare Life SciencesLegal20.020.00.0
Harvey's Legal Agent Benchmark - ImmigrationLegal0.00.00.0
Harvey's Legal Agent Benchmark - International Trade SanctionsLegal0.00.00.0
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

Gemini 3.7 FlashvsMuse Wondernormalised score marginsData Nov 15, 2026Data: Vals AI, Arena, Center for AI Safety and 25 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…