Atlas

Compare Models

Claude Sonnet 5.5 vs Muse Wonder

Claude Sonnet 5.5 leads by 3.0 capability points (≈95% interval +0.6 to +5.5): a clear lead over Muse Wonder · P(higher) ≈ 99%

Wins 104 of 145 shared benchmarks (40 lost, 1 tied)

Capability vs. relative cost
Claude Sonnet 5.5Muse 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.

Claude Sonnet 5.5vsMuse Wondercapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins104 – 40 · 145 shared, 1 tied

Comparing matching setups only (8 benchmarks excluded) · suites counted once
← Claude Sonnet 5.5Muse Wonder →

Claude Sonnet 5.5vsMuse Wonderwins by categoryData Nov 15, 2026Data: Artificial Analysis, Vals AI, Mercor and 43 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Muse Wonder minus Claude Sonnet 5.5.

LisanBench Difficulty-WeightedGames29838303+57.0
AA-PersonalityBench Transparency indexChat & Writing10056.2−44.2
LisanBench Path LengthGames1378727973+44.1
DeepResearch Bench II (Mercor 130-task set)Search51.710.8−40.9
HoneyBench v0.2 (hack rate)Safety41.681.0−39.4
Code2Video-AA EloCode1443997−36.2
ChessBench CoherenceGames64.6100.0+35.4
FrontierML ($200 wallet)Agents50.419.3−31.2
Terminal-Bench-Science 0.1Science53.325.7−27.6
AA-PersonalityBench Warmth indexChat & Writing73.3100+26.9
AA-PersonalityBench Slop indexChat & Writing27.754.3+26.9
LMArena Agent Arena Net Improvement (Steering Burden)Agents16.01.8−26.8
OpenWorldEngineer (Mean@8)Code39.012.7−26.3
sudo L7Code41.715.6−26.1
Terminal-Bench Science 0.1 (Vals)Science47.122.9−24.3
MysteryMechanismScience49.127.5−21.6
APEX-1 (Mercor, mean score)Professional Work53.975.1+21.2
AA-PersonalityBench Conviction indexChat & Writing58.179.1+21.2
HardGeoBench Avg DistanceMultimodal6831369−21.1
Furniture Assembly BenchmarkMultimodal75.054.2−20.8
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine41.722.2−19.4
Bug Hunt BenchCode48.930.2−18.7
DAYJOB: Data ScienceProfessional Work35.316.7−18.6
HANDBOOK.md Agents v1.1Professional Work47.729.5−18.2
Visa Agentic Commerce BenchmarkAgents64.482.2+17.8
RiemannBench (November 2026 rescoring)Math76.859.2−17.6
FrontierCode 2.0 - Pass RateCode46.428.8−17.6
FrontierCode 2.0 - ScoreCode42.225.3−16.9
Decision-Bench v1.1Professional Work51.234.6−16.6
LMArena Code Arena WebDevCode17731649−16.2
LMArena Agent Arena Praise vs ComplaintAgents30.5-1.8−16.2
IOI (Vals v2)Code83.166.9−16.1
ProofBench v1.1Math100.084.0−16.0
AA-PersonalityBench Expressiveness indexChat & Writing73.388.6+15.4
Code Migration v1.1Code86.972.4−14.5
Humanity's Last Exam DiamondGeneral QA53.339.0−14.3
Harvey's Legal Agent BenchmarkLegal2.917.1+14.2
Mystery Game PuzzlesGames65.051.0−14.0
GameBench 2Games91.477.8−13.6
HLAB-Vals-1.1Legal10.423.8+13.3
Conceptual Reasoning IndexIndexes77.167.7−12.8
SWE-bench Multimodal (Vals v1, 480 Tasks)Code58.345.6−12.7
APEX-SWECode66.454.6−11.8
LMCAGeneral QA68.356.8−11.4
DistressBenchSafety74.685.9+11.3
HardGeoBench <=100miMultimodal50.038.8−11.2
ChessBench EloGames8571046+11.2
Artificial Analysis Intelligence Index v5.0.1Indexes54.543.6−11.0
Artificial Analysis Intelligence Index v5.0Indexes54.243.3−10.9
Vending-Bench 2Games111659432−10.5
DeepFinanceFinance & Accounting54.644.2−10.4
FrontierMath Tier 4 (v2) - PrivateMath80.570.7−9.8
AA-PersonalityBench Informativeness indexChat & Writing84.894.3+9.6
DAYJOB: LawLegal27.036.6+9.6
VIALSMultimodal41.031.5−9.5
ACCoRDGeneral QA85.776.2−9.5
GBENCH One-shot Coding GScoreGames67.458.1−9.3
Vibe Code Bench v1.1Code92.483.1−9.3
EnterpriseBench: CoreCraftProfessional Work70.861.5−9.3
Artificial Analysis Engineering IndexIndexes58.549.6−8.9
Real-SWE 1.1Code51.242.4−8.8
HardGeoBench <=10miMultimodal36.327.5−8.7
AA-Briefcase v1.1 EloAgents18161659−8.6
Excel Modeling BenchmarkFinance & Accounting75.767.1−8.6
Agents' Last Exam v1.1 ScoreAgents60.468.4+8.0
MineBench — Bradley-Terry RatingGames19981868−7.9
Code MigrationCode69.862.6−7.3
LMArena Agent Arena Confirmed SuccessAgents19.55.4−7.1
LMArena Text Arena EloChat & Writing14721511+6.7
APEX-Agents 1.1Professional Work83.777.0−6.7
FrontierMath Tiers 1-3 (v2) - PrivateMath88.882.1−6.7
SpeciEval (Pooled English Runs)Safety91.097.5+6.5
HardGeoBench Country AccuracyMultimodal58.852.5−6.2
Ramp Accounting BenchFinance & Accounting42.548.7+6.2
ChessBench AccuracyGames75.870.2−6.0
LMArena Agent Arena Steering BurdenAgents14.52.6−6.0
micro1 PharmaBenchHealth & Medicine11.417.2+5.8
AA-PersonalityBench Imagination indexChat & Writing84.890.5+5.8
AA-PersonalityBench Organization indexChat & Writing84.890.5+5.8
Terminal-Bench 4.0 (Artificial Analysis)Code43.938.4−5.6
Artificial Analysis Healthcare & Medical IndexIndexes54.248.7−5.5
OncologyBench Strict (25 tasks, October 9 rubric)Health & Medicine29.324.0−5.3
Artificial Analysis Legal IndexIndexes52.246.9−5.3
Artificial Analysis Progress Index v1Indexes351336−5.2
CursorBench 4.0Code55.550.4−5.1
DAYJOB: TaxFinance & Accounting10.86.0−4.8
GDP.pdf (Artificial Analysis, All-pass)Professional Work25.830.6+4.8
Artificial Analysis Coding Agent Index v1.5Indexes68.463.7−4.7
Artificial Analysis Finance & Accounting IndexIndexes55.851.1−4.6
CharXiv Extended (Mercor held-out)Multimodal84.279.6−4.6
CritPt v2 (agentic, avg@5)Science75.771.1−4.6
Artificial Analysis Economics IndexIndexes54.149.5−4.5
MKA (Artificial Analysis)Professional Work52.047.5−4.5
DAYJOB: AccountingFinance & Accounting35.230.8−4.4
LiveBench 2.0Agents82.578.1−4.4
TaxCalcBench TY25Finance & Accounting18.022.0+4.0
LMArena Agent Arena Bash ReliabilityAgents12.74.8−4.0
AutomationBench-AA 1.0.6Agents71.868.1−3.7
Artificial Analysis Intelligence Index v4.3.2Indexes56.052.4−3.6
Artificial Analysis Strategy & Ops IndexIndexes58.154.5−3.6
Epoch Capabilities Index (ECI)Indexes166162−3.5
AA-Omniscience IndexIndexes32.325.5−3.4
Excel Modeling Benchmark v1.1Finance & Accounting75.972.5−3.4
CritPt (avg@5)Science31.428.3−3.1
Public Benefits Bench v1.1Professional Work67.270.2+3.0
BullshitBench v2 LeaderboardSafety60.057.0−3.0
DualEntry Accounting AI BenchmarkFinance & Accounting80.283.2+3.0
Eval Connections One-Shot with TrapsGames76.073.0−3.0
GDPval-AA v2.1Professional Work18381770−3.0
DAYJOB: HealthcareHealth & Medicine15.218.0+2.8
SWE Atlas v1.1 - OverallCode67.564.8−2.7
DAYJOB: FinanceFinance & Accounting24.021.3−2.7
APEX-Accounting 1.1Finance & Accounting64.567.1+2.6
RuneBenchGames4.85.0+2.5
DTBenchGeneral QA97.395.2−2.1
DAXBench v2Code86.284.0−2.1
τ³-Banking (Mercor run, bm25_grep retrieval)Agents54.051.9−2.1
WellnessBenchSafety57.055.0−2.0
Access to Justice BenchmarkLegal74.973.0−2.0
MissionReportProfessional Work63.862.1−1.7
Toloka Arena (v3)Agents64.065.7+1.7
NoCoT-Bench NCKI (kspine v3)General QA107109+1.6
CharXiv (descriptive and reasoning combined)Multimodal96.194.5−1.6
Humanity's Last ExamGeneral QA46.147.6+1.5
NoCoT-Bench NCRI 15.2General QA95.197.7+1.5
MMLU ProGeneral QA91.089.6−1.4
GDPxProfessional Work8.26.8−1.4
SWE-Marathon v1.1Code51.352.5+1.3
Financial Coaching BenchFinance & Accounting74.973.8−1.2
MazeBench ASCIIGames1.00.0−1.0
GDPval Extended (Mercor held-out rubric score)Professional Work70.970.0−0.9
MMMU ProMultimodal85.884.9−0.9
SpatialBench V1 Clean RerunsScience72.871.9−0.9
LMArena Agent Arena Tool HallucinationAgents0.9-0.5−0.7
ComplexConstraints (November 2026 rescoring)Chat & Writing67.267.9+0.7
BuseyBench SVGMultimodal6.56.6+0.7
SciCode 1.0.1 (Artificial Analysis)Code61.060.3−0.7
Immigration BenchLegal69.069.6+0.7
FrontierCode 1.1 Main (Artificial Analysis)Code45.846.5+0.6
AA-LCR Extended (Mercor held-out)Search73.874.4+0.6
OTIS Mock AIME 2024-2025Math100.099.4−0.6
GDPval (Mercor 213-task rubric score)Professional Work86.587.0+0.5
Prophet Arena Arena Score (Agentic, OpenCode)Forecasting85.785.4−0.3
Prophet Arena Arena Score (Fixed Context)Forecasting85.385.2−0.1
Dual-Arm Sim (task completion)Agents0.00.00.0

Claude Sonnet 5.5vsMuse Wondernormalised score marginsData Nov 15, 2026Data: Artificial Analysis, Vals AI, Mercor and 43 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…