Atlas

Compare Models

Muse Spark 1.3 vs Muse Wonder

Muse Wonder leads by 2.4 capability points (≈95% interval +0.2 to +4.6): a clear lead over Muse Spark 1.3 · P(higher) ≈ 98%

Wins 112 of 153 shared benchmarks (40 lost, 1 tied)

Capability vs. relative cost
Muse Spark 1.3Muse 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.

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

Benchmark wins40 – 112 · 153 shared, 1 tied

Comparing matching setups only (5 benchmarks excluded) · suites counted once
← Muse Spark 1.3Muse Wonder →

Muse Spark 1.3vsMuse Wonderwins by categoryData Nov 15, 2026Data: Artificial Analysis, Vals AI, Mercor and 41 othersVillage LLM Atlasatlas.village.dev

Margins

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

AA-PersonalityBench Organization indexChat & Writing63.890.5+26.9
ProofBench v1.1Math58.084.0+26.0
Mystery Game PuzzlesGames25.051.0+26.0
Terminal-Bench 4.0Code24.749.5+24.7
FrontierMath Tier 4 (v2) - PrivateMath46.370.7+24.4
CheatBenchSafety37.012.7+24.3
AA-PersonalityBench Slop indexChat & Writing75.354.3−21.2
Snorkel Agentic Coding 2.0Code25.043.6+18.6
VoxelBench Text RatingGames16161870+18.1
Visa Agentic Commerce BenchmarkAgents64.482.2+17.8
AA-PersonalityBench Expressiveness indexChat & Writing71.488.6+17.3
SnorkelManufacturingProfessional Work10.527.8+17.3
TaxCalcBench TY25Finance & Accounting6.022.0+16.0
LLM Creative Story WritingChat & Writing0.62.0+15.8
APEX-Agents 1.1Professional Work61.377.0+15.7
APEX-SWECode39.154.6+15.5
AA-PersonalityBench Transparency indexChat & Writing41.056.2+15.4
Code MigrationCode47.462.6+15.1
Terminal-Bench-Science 0.1Science11.025.7+14.8
Decision-Bench v1.1Professional Work19.834.6+14.8
TextQuests Game ProgressGames33.047.6+14.7
GDPval Extended (Mercor held-out rubric score)Professional Work55.970.0+14.1
DAYJOB: LawLegal22.836.6+13.8
Terminal-Bench Science 0.1 (Vals)Science10.022.9+12.9
DAYJOB: AccountingFinance & Accounting18.030.8+12.8
HiL-Bench (Human-in-Loop Benchmark)Agents66.979.7+12.8
Finance Agent v2 - General Qualitative AnalysisFinance & Accounting83.470.8−12.6
Eval Connections One-Shot with TrapsGames84.073.0−11.0
Finance Agent v2 - Financial ModelingFinance & Accounting33.522.5−11.0
DAYJOB: Data ScienceProfessional Work6.016.7+10.7
FrontierCS 2Code12.522.9+10.4
IOI (Vals v2)Code56.666.9+10.4
AutomationBench-AA 1.0.6Agents57.968.1+10.2
DAYJOB: HealthcareHealth & Medicine7.818.0+10.2
Snorkel Digital Work IndexIndexes16.326.4+10.1
MMMU-Pro Extended (Mercor held-out)Multimodal33.843.2+9.4
Artificial Analysis Coding Agent Index v1.5Indexes54.363.7+9.4
GameBench 2Games68.577.8+9.3
ChessBench EloGames12011046−9.2
Snorkel SWE-bench CLICode2.211.4+9.2
BuseyBench SVGMultimodal5.76.6+8.9
ChessBench CoherenceGames91.2100.0+8.8
CursorBench 4.0Code41.650.4+8.8
MysteryMechanismScience36.027.5−8.6
WeirdML v3Science6.715.2+8.5
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal15.023.0+8.0
Finance Agent v2 - Market AnalysisFinance & Accounting69.962.1−7.8
FrontierMath Tiers 1-3 (v2) - PrivateMath74.482.1+7.7
Artificial Analysis Coding Agent Index v2.1Indexes36.844.5+7.7
AA-PersonalityBench Imagination indexChat & Writing98.190.5−7.7
CyberBench v1.1 OverallCybersecurity72.780.4+7.6
AA-LCR Extended (Mercor held-out)Search66.974.4+7.5
PCT Political Inconsistency (CAIS)Safety40.633.3+7.4
MineBench — Bradley-Terry RatingGames17471868+7.3
DeepFinanceFinance & Accounting37.044.2+7.2
Artificial Analysis Intelligence Index v5.0.1Indexes36.643.6+7.0
Artificial Analysis Intelligence Index v5.0Indexes36.443.3+6.9
HackerRank-BenchCode42.949.7+6.9
Harvey's Legal Agent BenchmarkLegal23.817.1−6.7
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine15.622.2+6.7
DAYJOB: FinanceFinance & Accounting14.821.3+6.5
Humanity's Sixth SenseMultimodal37.443.7+6.3
Artificial Analysis Coding Agent Index v2.0Indexes45.752.0+6.3
SWE-Marathon v1.1Code46.352.5+6.3
Ramp Accounting BenchFinance & Accounting42.548.7+6.2
ChessBench AccuracyGames76.070.2−6.2
Design Arena Overall Code CategoriesCode13481385+6.1
RuneBenchGames4.65.0+6.1
WellnessBenchSafety61.055.0−6.0
FrontierCode 1.1 Main (Artificial Analysis)Code40.646.5+5.8
AA-PersonalityBench Informativeness indexChat & Writing10094.3−5.8
RiemannBench (November 2026 rescoring)Math53.659.2+5.6
SnorkelRevOpsProfessional Work12.618.0+5.4
GBENCH One-shot Coding GScoreGames52.758.1+5.4
Humanity's Last Exam DiamondGeneral QA33.739.0+5.3
Creative Writing v3 EloChat & Writing19072007+5.1
Terminal-Bench 4.0 (Artificial Analysis)Code33.338.4+5.1
SnorkelFinance 2.0Finance & Accounting10.015.0+5.0
Legal Research BenchLegal55.350.5−4.8
FrontierCode 2.0 (Artificial Analysis)Code23.127.9+4.8
ErdosBenchMath2.93.0+4.8
Immigration BenchLegal65.069.6+4.6
HANDBOOK.md Agents v1.1Professional Work33.929.5−4.4
Humanity's Last ExamGeneral QA43.247.6+4.4
Artificial Analysis Intelligence Index v4.3.2Indexes48.152.4+4.3
AA-Briefcase v1.1 EloAgents15811659+4.2
GDP.pdf (Artificial Analysis, All-pass)Professional Work26.630.6+4.0
Artificial Analysis Progress Index v1Indexes324336+4.0
GDPval-AA v2.1Professional Work16811770+3.9
AA-PersonalityBench Warmth indexChat & Writing96.2100+3.8
Finance Agent v2 - All PassFinance & Accounting49.545.9−3.6
Web Clone BenchCode0.23.7+3.5
SnorkelUnderwrite 2.0Finance & Accounting19.623.0+3.4
DeepSWE v1.1Code75.478.8+3.4
LiveBench 2.0Agents74.878.1+3.3
micro1 PharmaBenchHealth & Medicine13.917.2+3.3
FrontierML ($200 wallet)Agents16.019.3+3.3
Epoch Capabilities Index (ECI)Indexes158162+3.3
Professional Reasoning Benchmark - LegalLegal61.658.4−3.2
CritPt v2 (agentic, avg@5)Science68.071.1+3.1
Finance Agent v2 - PrecedentsFinance & Accounting35.232.1−3.1
ComplexConstraints (November 2026 rescoring)Chat & Writing64.867.9+3.1
LMCAGeneral QA53.956.8+3.0
Creative Writing v3 Rubric ScoreChat & Writing15.616.2+2.9
Vals Index v2.1Indexes58.261.1+2.9
Finance Agent v2 - Disclosure AnalysisFinance & Accounting70.167.3−2.8
Financial Coaching BenchFinance & Accounting71.073.8+2.8
Vibe Code Bench v1.1Code85.983.1−2.8
HoneyBench v0.2 (hack rate)Safety83.681.0+2.6
LMArena Text Arena EloChat & Writing14961511+2.6
SWE-bench Multimodal (Vals v1, 480 Tasks)Code43.145.6+2.5
DAYJOB: TaxFinance & Accounting3.66.0+2.4
NoCoT-Bench NCKI (kspine v3)General QA113109−2.3
CritPt (avg@5)Science26.028.3+2.3
CharXiv Extended (Mercor held-out)Multimodal81.879.6−2.2
APEX-Accounting 1.1Finance & Accounting64.967.1+2.2
Conceptual Reasoning IndexIndexes66.267.7+2.1
APEX-1 (Mercor, mean score)Professional Work73.075.1+2.1
AA-PersonalityBench Conviction indexChat & Writing77.279.1+1.9
SnorkelLegalLegal36.538.3+1.8
Code2Video-AA EloCode1019997−1.8
LMArena Agent Arena Praise vs ComplaintAgents1.7-1.8−1.8
sudo L7Code13.915.6+1.7
LMArena Agent Arena Bash ReliabilityAgents2.04.8+1.4
CUA-benchAgents5.87.2+1.3
DTBenchGeneral QA96.595.2−1.3
Tax Agent BenchFinance & Accounting72.471.2−1.3
SpeciEval (Pooled English Runs)Safety98.797.5−1.2
Harvey LAB-AA v1.1 (hallucination-gated)Legal8.910.0+1.1
ACCoRDGeneral QA75.176.2+1.1
DAXBench v2Code85.184.0−1.1
DistressBenchSafety84.985.9+1.0
Professional Reasoning Benchmark - FinanceFinance & Accounting59.558.7−0.9
LMArena Code Arena WebDevCode16551649−0.9
LMArena Agent Arena Confirmed SuccessAgents3.75.4+0.8
LMArena Agent Arena Net Improvement (Steering Burden)Agents2.21.8−0.8
Access to Justice BenchmarkLegal72.273.0+0.8
Finance Agent v2 - ComparablesFinance & Accounting47.247.9+0.8
LMArena Agent Arena Tool HallucinationAgents0.8-0.5−0.7
CharXiv (descriptive and reasoning combined)Multimodal93.994.5+0.6
SciCode 1.0.1 (Artificial Analysis)Code59.760.3+0.6
Bug Hunt BenchCode30.730.2−0.5
EnterpriseBench: CoreCraftProfessional Work61.961.5−0.4
Finance Agent v2 - Earnings AnalysisFinance & Accounting77.276.8−0.4
Finance Agent v2 - General Quantitative AnalysisFinance & Accounting78.077.6−0.4
Excel Modeling BenchmarkFinance & Accounting67.467.1−0.3
NoCoT-Bench NCRI 15.2General QA98.397.7−0.3
τ³-Banking (Mercor run, bm25_grep retrieval)Agents52.251.9−0.3
OTIS Mock AIME 2024-2025Math99.299.4+0.3
AA-Omniscience IndexIndexes25.025.5+0.2
Finance Agent v2 - AdjustmentsFinance & Accounting52.852.7−0.1
LMArena Agent Arena Steering BurdenAgents2.62.60.0
MazeBench ASCIIGames0.00.00.0

Muse Spark 1.3vsMuse Wondernormalised score marginsData Nov 15, 2026Data: Artificial Analysis, Vals AI, Mercor and 41 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…