Atlas

Compare Models

Muse Spark 1.3 vs Muse Wonder Full

Muse Wonder Full leads by 3.5 capability points (≈95% interval +1.2 to +5.8): a clear lead over Muse Spark 1.3 · P(higher) ≈ 100%

Wins 53 of 64 shared benchmarks (11 lost)

Capability vs. relative cost
Muse Spark 1.3Muse Wonder FullClosed 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.

Muse Spark 1.3vsMuse Wonder Fullcapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins11 – 53 · 64 shared

Comparing matching setups only (1 benchmark excluded) · suites counted once
← Muse Spark 1.3Muse Wonder Full →

Muse Spark 1.3vsMuse Wonder Fullwins by categoryData Nov 15, 2026Data: Vals AI, Artificial Analysis, Arena and 5 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Muse Wonder Full minus Muse Spark 1.3.

ProofBench v1.1Math58.094.0+36.0
Terminal-Bench 4.0Code24.758.1+33.3
Terminal-Bench Science 0.1 (Vals)Science10.037.1+27.1
Terminal-Bench-Science 0.1Science11.037.1+26.2
IOI (Vals v2)Code56.679.8+23.3
Artificial Analysis Coding Agent Index v2.1Indexes36.854.5+17.7
Humanity's Last Exam DiamondGeneral QA33.748.5+14.8
FrontierCode 2.0 (Artificial Analysis)Code23.137.8+14.6
Artificial Analysis Coding Agent Index v1.5Indexes54.368.0+13.7
Code MigrationCode47.461.0+13.6
AutomationBench-AA 1.0.6Agents57.971.3+13.5
Artificial Analysis Intelligence Index v5.0.1Indexes36.650.1+13.5
Artificial Analysis Intelligence Index v5.0Indexes36.449.7+13.3
Artificial Analysis Coding Agent Index v2.0Indexes45.758.9+13.2
LMArena Agent Arena Net Improvement (Steering Burden)Agents2.29.0+12.8
Finance Agent v2 - Financial ModelingFinance & Accounting33.523.7−9.8
RiemannBench (November 2026 rescoring)Math53.663.2+9.6
Bug Hunt BenchCode30.740.0+9.3
Finance Agent v2 - General Qualitative AnalysisFinance & Accounting83.474.2−9.2
LMArena Code Arena WebDevCode16551723+8.9
AA-Briefcase v1.1 EloAgents15811729+8.1
SRE Bench — Fully SolvedCode3.811.8+8.0
GDPval-AA v2.1Professional Work16811857+7.8
Artificial Analysis Intelligence Index v4.3.2Indexes48.155.8+7.7
CritPt v2 (agentic, avg@5)Science68.075.1+7.1
SWE-Marathon v1.1Code46.353.1+6.9
LMArena Agent Arena Praise vs ComplaintAgents1.715.4+6.9
CritPt (avg@5)Science26.032.9+6.9
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine15.622.2+6.7
Artificial Analysis Progress Index v1Indexes324343+6.6
Vals Index v2.1Indexes58.264.6+6.4
Immigration BenchLegal65.071.2+6.2
GDP.pdf (Artificial Analysis, All-pass)Professional Work26.632.6+6.0
Finance Agent v2 - Market AnalysisFinance & Accounting69.964.9−5.0
Humanity's Last Exam (Text-Only)General QA48.753.6+4.9
Vals RSI Index v1.1 (four tasks)Agents19.624.4+4.8
Excel Modeling BenchmarkFinance & Accounting67.471.9+4.5
FrontierCode 1.1 Main (Artificial Analysis)Code40.644.9+4.2
SWE-bench Multimodal (Vals v1, 480 Tasks)Code43.147.3+4.2
Harvey LAB-AA v1.1 (hallucination-gated)Legal8.913.1+4.2
Financial Coaching BenchFinance & Accounting71.075.0+4.0
Finance Agent v2 - AdjustmentsFinance & Accounting52.856.7+3.9
LMArena Agent Arena Confirmed SuccessAgents3.711.1+3.7
DeepSWE v1.1Code75.479.0+3.6
LMArena Agent Arena Steering BurdenAgents2.69.6+3.5
LMArena Agent Arena Bash ReliabilityAgents2.08.3+3.1
SciCode 1.0.1 (Artificial Analysis)Code59.762.6+2.9
AA-Omniscience IndexIndexes25.030.1+2.6
Tax Agent BenchFinance & Accounting72.475.0+2.6
Access to Justice BenchmarkLegal72.274.7+2.5
Harvey's Legal Agent BenchmarkLegal23.821.7−2.1
Finance Agent v2 - General Quantitative AnalysisFinance & Accounting78.080.1+2.1
ProgramBenchCode2.50.5−2.0
Legal Research BenchLegal55.353.4−1.9
Finance Agent v2 - Disclosure AnalysisFinance & Accounting70.171.6+1.5
Finance Agent v2 - Earnings AnalysisFinance & Accounting77.278.6+1.4
MysteryMechanismScience36.035.1−0.9
CyberBench v1.1 OverallCybersecurity72.773.6+0.9
AA-LCR v1.1Search83.083.7+0.7
Finance Agent v2 - All PassFinance & Accounting49.549.0−0.5
LMArena Agent Arena Tool HallucinationAgents0.80.0−0.4
Finance Agent v2 - ComparablesFinance & Accounting47.246.8−0.4
Finance Agent v2 - PrecedentsFinance & Accounting35.235.0−0.2
Vibe Code Bench v1.1Code85.986.0+0.1

Muse Spark 1.3vsMuse Wonder Fullnormalised score marginsData Nov 15, 2026Data: Vals AI, Artificial Analysis, Arena and 5 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…