Atlas

Compare Models

GPT-6.1 Sol vs Muse Wonder

GPT-6.1 Sol leads by 2.7 capability points (≈95% interval +0.3 to +5.2): a clear lead over Muse Wonder · P(higher) ≈ 99%

Wins 101 of 183 shared benchmarks (78 lost, 4 tied)

Capability vs. relative cost
GPT-6.1 SolMuse 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.

GPT-6.1 SolvsMuse Wondercapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins101 – 78 · 183 shared, 4 tied

Comparing matching setups only (58 benchmarks excluded) · suites counted once
← GPT-6.1 SolMuse Wonder →

GPT-6.1 SolvsMuse Wonderwins by categoryData Nov 15, 2026Data: Vals AI, Artificial Analysis, Center for AI Safety and 37 othersVillage LLM Atlasatlas.village.dev

Margins

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

AA-PersonalityBench Informativeness indexChat & Writing27.794.3+67.3
LisanBench Path LengthGames873627973+59.8
HPCT Refusal (CAIS)Safety42.097.0+55.0
VoxelBench Text RatingGames25501870−48.4
LisanBench Difficulty-WeightedGames40048303+46.0
VCT-Refusal (CAIS)Safety16.160.2+44.1
CyberBench v1.1 OverallCybersecurity39.380.4+41.1
AA-PersonalityBench Slop indexChat & Writing14.354.3+40.4
AA-PersonalityBench Expressiveness indexChat & Writing50.588.6+38.5
AA-PersonalityBench Transparency indexChat & Writing92.456.2−36.5
AA-PersonalityBench Warmth indexChat & Writing65.7100+34.6
Terminal-Bench-Science 0.1Science58.125.7−32.4
ChessBench EloGames15621046−30.7
CheatBenchSafety42.873.1−30.3
Code2Video-AA EloCode1369997−30.2
Terminal-Bench Science 0.1 (Vals)Science52.922.9−30.0
IOI (Vals v2)Code96.966.9−29.9
FrontierMath Tier 4 (v2) - PrivateMath100.070.7−29.3
TextQuests Average Cumulative HarmSafety29.320.8+29.1
Mystery Game PuzzlesGames80.051.0−29.0
AA-PersonalityBench Imagination indexChat & Writing61.990.5+28.8
OpenWorldEngineer (Mean@8)Code41.212.7−28.4
NoCoT-Bench NCRI 15.2General QA14997.7−28.2
Furniture Assembly BenchmarkMultimodal80.054.2−25.8
AA-PersonalityBench Conviction indexChat & Writing54.379.1+25.0
TextQuests Game ProgressGames71.447.6−23.8
EnigmaEvalGeneral QA59.536.1−23.4
WeirdML v3Science38.315.2−23.1
EBR-bench v4 (card ban, single agent)Games54.331.4−22.9
DataCenterBench Tier 1Agents38.316.7−21.7
HoneyBench v0.2 (hack rate)Safety59.581.0−21.5
GBENCH One-shot Coding GScoreGames79.058.1−21.0
Vals Science IndexIndexes59.639.1−20.5
LMArena Agent Arena Net Improvement (Steering Burden)Agents12.51.8−20.2
RiemannBench (November 2026 rescoring)Math79.259.2−20.0
Visa Agentic Commerce BenchmarkAgents62.282.2+20.0
Humanity's Last Exam DiamondGeneral QA58.839.0−19.8
MysteryMechanismScience46.427.5−18.9
MMMU-Pro Extended (Mercor held-out)Multimodal61.343.2−18.1
sudo L7 (October 2026 rescoring)Code33.916.1−17.8
BioSecBench-Refusal V1 Clean RerunsSafety37.655.3+17.7
VIALSMultimodal49.131.5−17.6
SnorkelLegalLegal20.738.3+17.6
HLAB-Vals-1.1Legal6.323.8+17.5
HiL-Bench (Human-in-Loop Benchmark)Agents62.379.7+17.3
CrossSearch (pass^3)Search59.042.0−17.0
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine38.322.2−16.1
TaxCalcBench TY25Finance & Accounting38.022.0−16.0
ProofBench v1.1Math99.084.0−15.0
LMArena Agent Arena Praise vs ComplaintAgents27.5-1.8−14.7
Decision-Bench v1.1Professional Work49.134.6−14.5
Machiavelli Deception (CAIS)Safety78.883.6−14.4
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal8.823.0+14.2
GameBench 2Games92.077.8−14.2
NoCoT-Bench NCKI (kspine v3)General QA130109−14.0
EQ-Bench 4Chat & Writing12721325+13.8
BuseyBench SVGMultimodal8.06.6−13.5
MASKSafety98.384.9−13.4
Terminal-Bench 4.0 (Artificial Analysis)Code51.538.4−13.1
FrontierCode 2.0 (Artificial Analysis)Code40.627.9−12.8
BioMysteryBench (Vals Terminus 2)Science79.667.0−12.6
ChessBench CoherenceGames87.5100.0+12.5
Professional Reasoning Benchmark - FinanceFinance & Accounting46.258.7+12.5
Legal Research BenchLegal38.550.5+12.0
DistressBenchSafety74.285.9+11.7
Harvey's Legal Agent BenchmarkLegal5.417.1+11.7
FrontierMath Tiers 1-3 (v2) - PrivateMath93.782.1−11.6
CritPt v2 (agentic, avg@5)Science82.671.1−11.4
ChessBench AccuracyGames80.770.2−11.3
EnterpriseBench: CoreCraftProfessional Work50.361.5+11.2
Artificial Analysis Intelligence Index v5.0Indexes54.543.3−11.2
Artificial Analysis Intelligence Index v5.0.1Indexes54.643.6−11.0
WellnessBenchSafety66.055.0−11.0
Public Benefits Bench v1.1Professional Work59.370.2+10.9
Professional Reasoning Benchmark - LegalLegal47.958.4+10.5
DAYJOB: HealthcareHealth & Medicine7.618.0+10.4
Artificial Analysis Engineering IndexIndexes59.449.6−9.8
AA-PersonalityBench Organization indexChat & Writing81.090.5+9.6
τ³-Banking (Mercor run, bm25_grep retrieval)Agents43.051.9+8.9
Tax Agent BenchFinance & Accounting62.371.2+8.9
Finance Agent v2 - Disclosure AnalysisFinance & Accounting58.467.3+8.8
DAXBench v2Code92.684.0−8.5
SnorkelFinance 2.0Finance & Accounting6.615.0+8.4
SnorkelManufacturingProfessional Work19.627.8+8.2
AA-Omniscience IndexIndexes41.525.5−8.0
BullshitBench v2 LeaderboardSafety65.057.0−8.0
Finance Agent v2 - AdjustmentsFinance & Accounting44.752.7+8.0
APEX-SWECode46.954.6+7.7
Finance Agent v2 - General Qualitative AnalysisFinance & Accounting63.270.8+7.6
HardGeoBench <=10miMultimodal35.027.5−7.5
LiveBench 2.0Agents85.478.1−7.3
GDPval-AA v2.1Professional Work16071770+7.2
Access to Justice BenchmarkLegal65.873.0+7.1
Conceptual Reasoning IndexIndexes72.767.7−6.8
Artificial Analysis Legal IndexIndexes53.746.9−6.8
Artificial Analysis Economics IndexIndexes56.349.5−6.8
SpatialViz-BenchMultimodal90.383.5−6.8
Vals Legal IndexIndexes68.775.3+6.5
MacroscopeBenchCode70.664.2−6.4
Finance Agent v2 - PrecedentsFinance & Accounting25.832.1+6.3
Finance Agent v2 - All PassFinance & Accounting39.945.9+6.0
LMCAGeneral QA62.856.8−5.9
SnorkelRevOpsProfessional Work12.118.0+5.9
Vibe Code Bench v1.1Code88.983.1−5.8
AA-Briefcase v1.1 EloAgents15551659+5.7
DAYJOB: AccountingFinance & Accounting25.230.8+5.6
Terminal-Bench 4.0Code55.149.5−5.6
Epoch Capabilities Index (ECI)Indexes169162−5.5
CharXiv Extended (Mercor held-out)Multimodal85.179.6−5.5
LMArena Agent Arena Steering BurdenAgents13.42.6−5.4
Vals Finance IndexIndexes55.861.2+5.4
Finance Agent v2 - ComparablesFinance & Accounting42.647.9+5.3
Finance Agent v2 - Earnings AnalysisFinance & Accounting71.576.8+5.3
Financial Coaching BenchFinance & Accounting68.573.8+5.2
Snorkel Digital Work IndexIndexes21.226.4+5.2
Artificial Analysis Finance & Accounting IndexIndexes56.251.1−5.1
LMArena Text Arena EloChat & Writing14821511+5.0
HardGeoBench <=100miMultimodal43.838.8−5.0
Code Migration v1.1Code77.372.4−5.0
DAYJOB: LawLegal31.836.6+4.8
ACCoRDGeneral QA80.976.2−4.7
SpatialBench V1 Clean RerunsScience67.371.9+4.6
SciCode 1.0.1 (Artificial Analysis)Code55.860.3+4.5
Artificial Analysis Progress Index v1Indexes348336−4.5
HardGeoBench Avg DistanceMultimodal12391369−4.0
PCT Political Inconsistency (CAIS)Safety37.333.3+4.0
APEX-Agents 1.1Professional Work73.277.0+3.8
Finance Agent v2.1Finance & Accounting61.865.5+3.7
Excel Modeling BenchmarkFinance & Accounting70.867.1−3.7
Humanity's Last ExamGeneral QA51.347.6−3.7
CritPt (avg@5)Science31.728.3−3.4
Artificial Analysis Healthcare & Medical IndexIndexes52.048.7−3.3
DAYJOB: FinanceFinance & Accounting18.021.3+3.3
LMArena Agent Arena Confirmed SuccessAgents11.75.4−3.2
Artificial Analysis Strategy & Ops IndexIndexes57.654.5−3.1
Finance Agent v2 - General Quantitative AnalysisFinance & Accounting74.577.6+3.1
ERQAMultimodal78.275.3−2.9
Humanity's Sixth SenseMultimodal46.643.7−2.9
APEX-Accounting 1.1Finance & Accounting64.367.1+2.8
FrontierCode 1.1 Main (Artificial Analysis)Code49.146.5−2.6
Code MigrationCode65.162.6−2.6
Snorkel WorkplaceAgentsProfessional Work15.017.4+2.4
Finance Agent v2 - Market AnalysisFinance & Accounting64.562.1−2.4
Ramp Accounting BenchFinance & Accounting46.448.7+2.3
ComplexConstraints (November 2026 rescoring)Chat & Writing70.167.9−2.2
IntPhys 2Multimodal72.370.2−2.1
SWE-bench Multimodal (Vals v1, 480 Tasks)Code43.545.6+2.1
DAYJOB: Data ScienceProfessional Work18.716.7−2.0
DAYJOB: TaxFinance & Accounting4.06.0+2.0
DualEntry Accounting AI BenchmarkFinance & Accounting81.283.2+2.0
SnorkelUnderwrite 2.0Finance & Accounting25.023.0−2.0
MindCube TinyMultimodal89.191.1+1.9
Immigration BenchLegal67.869.6+1.8
GDPxProfessional Work8.46.8−1.6
Snorkel SWE-bench CLICode9.811.4+1.6
AutomationBench-AA 1.0.6Agents66.668.1+1.5
CharXiv (descriptive and reasoning combined)Multimodal96.094.5−1.5
Excel Modeling Benchmark v1.1Finance & Accounting71.172.5+1.4
MMLU ProGeneral QA91.089.6−1.4
GDP.pdf (Artificial Analysis, All-pass)Professional Work32.030.6−1.4
DTBenchGeneral QA96.595.2−1.3
SciCode (main problems)Science14.615.8+1.2
SWE Atlas v1.1 - OverallCode65.964.8−1.1
MissionReportProfessional Work61.062.1+1.1
Vals Index v2.2Indexes63.964.9+0.9
HANDBOOK.md Agents v1.1Professional Work28.629.5+0.9
Snorkel Agentic Coding 2.0Code42.843.6+0.8
LMArena Agent Arena Tool HallucinationAgents1.0-0.5−0.8
Prophet Arena Arena Score (Agentic, OpenCode)Forecasting86.085.4−0.6
Finance Agent v2 - Financial ModelingFinance & Accounting23.122.5−0.6
Artificial Analysis Intelligence Index v4.3.2Indexes51.852.4+0.6
OTIS Mock AIME 2024-2025Math100.099.4−0.6
Prophet Arena Arena Score (Fixed Context)Forecasting84.885.2+0.3
Toloka Arena (v3)Agents66.065.7−0.3
MKA (Artificial Analysis)Professional Work47.747.5−0.2
MMMU ProMultimodal85.184.9−0.2
SpeciEval (Pooled English Runs)Safety97.497.5+0.1
Vals Index v2.1Indexes61.261.1−0.1
LMArena Agent Arena Bash ReliabilityAgents5.04.8−0.1
Dual-Arm Sim (task completion)Agents0.00.00.0
FrontierMemory v1.1Agents51.151.10.0
HardGeoBench Country AccuracyMultimodal52.552.50.0
OncologyBench Strict (25 tasks, October 9 rubric)Health & Medicine24.024.00.0

GPT-6.1 SolvsMuse Wondernormalised score marginsData Nov 15, 2026Data: Vals AI, Artificial Analysis, Center for AI Safety and 37 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…