Atlas

Compare Models

Claude Haiku 5.5 vs Muse Wonder

Muse Wonder leads by 2.2 capability points (≈95% interval +0.0 to +4.4): a clear lead over Claude Haiku 5.5 · P(higher) ≈ 98%

Wins 122 of 147 shared benchmarks (24 lost, 1 tied)

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

Benchmark wins24 – 122 · 147 shared, 1 tied

Comparing matching setups only (22 benchmarks excluded) · suites counted once
← Claude Haiku 5.5Muse Wonder →

Claude Haiku 5.5vsMuse Wonderwins by categoryData Nov 15, 2026Data: Artificial Analysis, Vals AI, Mercor and 33 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 Haiku 5.5.

LisanBench Difficulty-WeightedGames27118303+59.9
AA-PersonalityBench Organization indexChat & Writing41.090.5+50.0
LisanBench Path LengthGames1289727973+46.8
Toloka Arena (v3)Agents21.065.7+44.7
AA-PersonalityBench Slop indexChat & Writing10.554.3+44.2
Vending-Bench 2Games36029432+35.3
AutomationBench-AA 1.0.6Agents36.068.1+32.1
AA-PersonalityBench Warmth indexChat & Writing69.5100+30.8
AA-PersonalityBench Imagination indexChat & Writing63.890.5+26.9
FrontierMath Tier 4 (v2) - PrivateMath46.370.7+24.4
DeepResearch Bench II (Mercor 130-task set)Search34.310.8−23.5
Artificial Analysis Coding Agent Index v1.5Indexes41.463.7+22.2
AA-PersonalityBench Transparency indexChat & Writing77.256.2−21.2
Mystery Game PuzzlesGames30.051.0+21.0
BuseyBench SVGMultimodal4.66.6+20.4
TaxCalcBench TY25Finance & Accounting2.022.0+20.0
Code Migration v1.1Code52.472.4+20.0
IOI (Vals v2)Code47.366.9+19.7
HLAB-Vals-1.1Legal4.223.8+19.6
ChessBench CoherenceGames80.7100.0+19.3
AA-PersonalityBench Informativeness indexChat & Writing75.394.3+19.2
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal4.923.0+18.1
Code MigrationCode44.662.6+18.0
MineBench — Bradley-Terry RatingGames15821868+17.3
Terminal-Bench 4.0 (Artificial Analysis)Code21.738.4+16.7
NoCoT-Bench NCKI (kspine v3)General QA85.9109+16.3
NoCoT-Bench NCRI 15.2General QA68.397.7+16.1
Code2Video-AA EloCode1192997−15.9
Harvey's Legal Agent BenchmarkLegal1.317.1+15.8
FrontierML ($200 wallet)Agents34.919.3−15.7
LMArena Text Arena EloChat & Writing14221511+15.5
Artificial Analysis Strategy & Ops IndexIndexes39.154.5+15.4
DistressBenchSafety71.785.9+14.2
Terminal-Bench 4.0Code36.449.5+13.1
Terminal-Bench Science 0.1 (Vals)Science10.022.9+12.9
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine35.022.2−12.8
VoxelBench Text RatingGames16961870+12.4
EBR-bench v4 (card ban, single agent)Games43.831.4−12.4
AA-PersonalityBench Conviction indexChat & Writing67.679.1+11.5
AA-PersonalityBench Expressiveness indexChat & Writing77.288.6+11.5
MMMU-Pro Extended (Mercor held-out)Multimodal31.943.2+11.3
Artificial Analysis Healthcare & Medical IndexIndexes37.548.7+11.2
AA-LCR Extended (Mercor held-out)Search63.874.4+10.6
SpeciEval (Pooled English Runs)Safety86.997.5+10.6
APEX-1 (Mercor, mean score)Professional Work64.775.1+10.4
ChessBench EloGames8751046+10.2
GDP.pdf (Artificial Analysis, All-pass)Professional Work20.830.6+9.8
Bug Hunt BenchCode20.530.2+9.7
Humanity's Last Exam DiamondGeneral QA29.639.0+9.4
Decision-Bench v1.1Professional Work25.334.6+9.3
Artificial Analysis Intelligence Index v4.3.2Indexes43.452.4+9.0
Tax Agent BenchFinance & Accounting62.271.2+9.0
Finance Agent v2 - General Qualitative AnalysisFinance & Accounting61.970.8+8.9
τ³-Banking (Mercor run, bm25_grep retrieval)Agents43.051.9+8.9
APEX-Agents 1.1Professional Work68.377.0+8.7
Artificial Analysis Legal IndexIndexes38.746.9+8.2
Excel Modeling BenchmarkFinance & Accounting59.067.1+8.1
Immigration BenchLegal61.869.6+7.8
MysteryMechanismScience35.127.5−7.7
Artificial Analysis Finance & Accounting IndexIndexes43.651.1+7.6
Humanity's Last Exam (Text-Only)General QA42.750.1+7.4
AA-Omniscience IndexIndexes10.725.5+7.4
Vibe Code Bench v1.1Code90.483.1−7.4
LMArena Code Arena WebDevCode15941649+7.2
CritPt v2 (agentic, avg@5)Science64.071.1+7.1
Artificial Analysis Intelligence Index v5.0Indexes36.243.3+7.1
Artificial Analysis Intelligence Index v5.0.1Indexes36.543.6+7.1
Public Benefits Bench v1.1Professional Work63.270.2+7.0
Excel Modeling Benchmark v1.1Finance & Accounting65.572.5+7.0
FrontierMath Tiers 1-3 (v2) - PrivateMath75.182.1+7.0
Artificial Analysis Engineering IndexIndexes42.549.6+7.0
DualEntry Accounting AI BenchmarkFinance & Accounting76.283.2+7.0
Artificial Analysis Economics IndexIndexes42.749.5+6.8
Legal Research BenchLegal43.850.5+6.7
Furniture Assembly BenchmarkMultimodal47.554.2+6.7
Vals Index v2.1Indexes54.461.1+6.6
MKA (Artificial Analysis)Professional Work41.047.5+6.5
Ramp Accounting BenchFinance & Accounting42.348.7+6.4
HardGeoBench <=100miMultimodal45.038.8−6.2
GDPval-AA v2.1Professional Work16291770+6.2
VIALS-AA (pass^3)Multimodal49.155.3+6.2
Vals Index v2.2Indexes58.764.9+6.1
Eval Connections One-Shot with TrapsGames67.073.0+6.0
Finance Agent v2 - Financial ModelingFinance & Accounting28.322.5−5.8
MMMU ProMultimodal79.184.9+5.8
LMCAGeneral QA51.156.8+5.8
Terminal-Bench-Science 0.1Science20.025.7+5.7
CritPt (avg@5)Science22.628.3+5.7
SWE Atlas v1.1 - OverallCode59.464.8+5.4
ACCoRDGeneral QA81.576.2−5.4
SciCode 1.0.1 (Artificial Analysis)Code55.060.3+5.3
Finance Agent v2 - PrecedentsFinance & Accounting27.032.1+5.1
ProofBench v1.1Math89.084.0−5.0
Vals Legal IndexIndexes70.375.3+5.0
APEX-Accounting 1.1Finance & Accounting62.267.1+4.9
CyberBench v1.1 OverallCybersecurity75.580.4+4.9
AA-Briefcase v1.1 EloAgents15731659+4.7
FrontierCode 1.1 Main (Artificial Analysis)Code41.846.5+4.7
VIALS-AA (average of 3)Multimodal60.564.8+4.3
DeepFinanceFinance & Accounting40.044.2+4.2
GDPval Extended (Mercor held-out rubric score)Professional Work66.070.0+4.0
Vals Finance IndexIndexes57.461.2+3.8
FrontierCode 2.0 - ScoreCode28.925.3−3.6
ChessBench AccuracyGames66.970.2+3.5
Artificial Analysis Progress Index v1Indexes326336+3.4
Conceptual Reasoning IndexIndexes65.367.7+3.4
BioMysteryBench (Vals Terminus 2)Science70.467.0−3.3
CharXiv Extended (Mercor held-out)Multimodal82.979.6−3.3
Finance Agent v2 - Earnings AnalysisFinance & Accounting73.576.8+3.3
Finance Agent v2 - General Quantitative AnalysisFinance & Accounting74.477.6+3.2
Finance Agent v2 - AdjustmentsFinance & Accounting49.652.7+3.1
FrontierCode 2.0 - Pass RateCode31.928.8−3.1
Finance Agent v2 - All PassFinance & Accounting42.945.9+3.1
Finance Agent v2 - Disclosure AnalysisFinance & Accounting64.367.3+3.0
GameBench 2Games75.177.8+2.7
Finance Agent v2 - ComparablesFinance & Accounting45.347.9+2.6
SWE-Marathon v1.1Code50.052.5+2.5
HardGeoBench <=10miMultimodal30.027.5−2.5
VIALS-AA (pass@3)Multimodal71.473.9+2.5
LiveBench 2.0Agents75.678.1+2.5
LMArena Agent Arena Net Improvement (Steering Burden)Agents0.51.8+2.5
GBENCH One-shot Coding GScoreGames55.658.1+2.4
MMLU ProGeneral QA87.489.6+2.2
DAXBench v2Code81.984.0+2.1
CursorBench 4.0Code48.450.4+2.0
LMArena Agent Arena Confirmed SuccessAgents1.65.4+1.9
GDPxProfessional Work4.96.8+1.9
SWE-bench Multimodal (Vals v1, 480 Tasks)Code44.045.6+1.7
HardGeoBench Avg DistanceMultimodal13191369−1.5
GDPval (Mercor 213-task rubric score)Professional Work85.587.0+1.5
Epoch Capabilities Index (ECI)Indexes160162+1.4
LMArena Agent Arena Bash ReliabilityAgents7.44.8−1.3
HardGeoBench Country AccuracyMultimodal51.352.5+1.3
MazeBench ASCIIGames1.00.0−1.0
Finance Agent v2 - Market AnalysisFinance & Accounting63.162.1−1.0
LMArena Agent Arena Praise vs ComplaintAgents-3.5-1.8+0.8
LMArena Agent Arena Tool HallucinationAgents-2.1-0.5+0.8
Vals Science IndexIndexes38.539.1+0.6
OTIS Mock AIME 2024-2025Math98.999.4+0.6
Prophet Arena Arena Score (Agentic, OpenCode)Forecasting85.985.4−0.5
APEX-SWECode55.154.6−0.5
CharXiv (descriptive and reasoning combined)Multimodal94.194.5+0.4
DTBenchGeneral QA95.095.2+0.2
Prophet Arena Arena Score (Fixed Context)Forecasting85.485.2−0.2
Finance Agent v2.1Finance & Accounting65.465.5+0.2
LMArena Agent Arena Steering BurdenAgents2.42.6+0.1
BullshitBench v2 LeaderboardSafety57.057.00.0

Claude Haiku 5.5vsMuse Wondernormalised score marginsData Nov 15, 2026Data: Artificial Analysis, Vals AI, Mercor and 33 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…