Atlas

Compare Models

Muse Ember vs Muse Spark 1.3

Muse Ember leads by 1.1 capability points (≈95% interval −1.1 to +3.4): statistically tied · P(higher) ≈ 84%

Muse Spark 1.3 wins 30 of 51 shared benchmarks (20 lost, 1 tied)

Capability vs. relative cost
Muse EmberMuse Spark 1.3Closed weightsOpen weights

Benchmark wins20 – 30 · 51 shared, 1 tied

Comparing matching setups only · suites counted once
← Muse EmberMuse Spark 1.3 →
Data: Artificial Analysis, Surge AI, Arena and 11 othersatlas.village.dev

Margins

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

DAXBench v2Software › Data & ML62.885.1+22.3
Terminal-Bench 4.0Software › Software engineering39.424.7−14.6
APEX-Agents 1.1Professional › General knowledge work75.761.3−14.4
APEX-SWESoftware › Software engineering52.939.1−13.8
AutomationBench-AA 1.0.6Tool & computer use › Customer service & operations68.157.9−10.2
DrugDiscoveryBenchScience › Life sciences52.462.2+9.8
SimpleBenchKnowledge & reasoning › General knowledge72.781.8+9.1
Terminal-Bench-Science 0.1Science › Multidisciplinary20.011.0−9.0
Humanity's Last Exam DiamondKnowledge & reasoning › General knowledge25.133.7+8.6
AA-Omniscience v1.1 IndexKnowledge & reasoning › General knowledge7.723.8+8.0
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Professional › Legal22.715.0−7.7
Terminal-Bench 4.0 (Artificial Analysis)Software › Software engineering40.933.3−7.6
FrontierCUA v1 pass@1Tool & computer use › Computer & browser use7.50.0−7.5
GDP.xlsxProfessional › General knowledge work15.422.6+7.2
HiL-Bench (Human-in-Loop Benchmark)Tool & computer use › General agents74.066.9−7.1
Legal Research BenchProfessional › Legal49.055.3+6.2
LMArena Code Arena WebDevSoftware › Frontend & design16081654+6.1
ComplexConstraints (November 2026 rescoring)Language & writing › Instruction following58.864.8+6.0
Professional Reasoning Benchmark - LegalProfessional › Legal55.661.6+6.0
CritPt v2 (agentic, avg@5)Science › Physical sciences62.368.0+5.7
Chartography (November 2026 rescoring)Perception & media › Documents & charts79.384.6+5.3
DAYJOB: Healthcare (October 2026 rescoring)Health & medicine17.222.0+4.8
HANDBOOK.md Agents v1.1Tool & computer use › Customer service & operations29.833.9+4.1
FrontierCode 2.0 (Artificial Analysis)Software › Software engineering27.223.1−4.1
KradleBenchGames25.721.7−4.0
TaxCalcBench TY25Professional › Finance, accounting & tax2.06.0+4.0
GDP.pdf (Artificial Analysis, All-pass)Professional › General knowledge work22.826.6+3.8
GDP.pdf (November 2026 rescoring)Professional › General knowledge work49.253.0+3.8
Professional Reasoning Benchmark - FinanceProfessional › Finance, accounting & tax56.059.5+3.6
AA-Briefcase v1.1 EloProfessional › General knowledge work15201581+3.3
BuseyBench SVGPerception & media › Generation5.45.7+3.1
Artificial Analysis Coding Agent Index v2.1Software › Software engineering39.736.8−2.9
DAYJOB: Finance (October 2026 rescoring)Professional › Finance, accounting & tax24.227.0+2.8
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & medicine18.315.6−2.8
LMArena Agent Arena Bash ReliabilityTool & computer use › General agents6.61.2−2.7
GBENCH One-shot Coding GScoreGames56.453.7−2.7
APEX-1 (Mercor, mean score)Professional › General knowledge work70.473.0+2.6
LMArena Agent Arena Confirmed SuccessTool & computer use › General agents-2.02.5+2.2
Vibe Code Bench v1.1Software › Frontend & design83.785.9+2.2
LMArena Agent Arena Steering BurdenTool & computer use › General agents4.80.8−2.0
Artificial Analysis Progress Index v1Composite indexes329324−1.5
LMArena Agent Arena Praise vs ComplaintTool & computer use › General agents-2.70.1+1.4
τ³-Banking (Mercor run, bm25_grep retrieval)Tool & computer use › Customer service & operations53.352.2−1.1
Tax Agent BenchProfessional › Finance, accounting & tax71.472.4+1.1
Artificial Analysis Intelligence Index v5.0.1Composite indexes36.036.6+0.6
DeepFinanceProfessional › Finance, accounting & tax36.537.0+0.5
LMArena Agent Arena Net Improvement (Steering Burden)Tool & computer use › General agents1.01.1+0.3
Artificial Analysis Coding Agent Index v2.0Software › Software engineering45.945.7−0.2
EnterpriseBench: CoreCraftTool & computer use › Customer service & operations62.161.9−0.2
LMArena Agent Arena Tool HallucinationTool & computer use › Tool calling0.91.0+0.1
RiemannBench (November 2026 rescoring)Math53.653.60.0
Data: Artificial Analysis, Surge AI, Arena and 11 othersatlas.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…