Atlas

Compare Models

Claude Haiku 5.5 vs Muse Ember

Muse Ember leads by 0.9 capability points (≈95% interval −1.3 to +3.2): statistically tied · P(higher) ≈ 79%

Wins 29 of 51 shared benchmarks (21 lost, 1 tied)

Capability vs. relative cost
Claude Haiku 5.5Muse EmberClosed weightsOpen weights

Benchmark wins21 – 29 · 51 shared, 1 tied

Comparing matching setups only (1 benchmark excluded) · suites counted once
← Claude Haiku 5.5Muse Ember →
Data: Artificial Analysis, Vals AI, Arena and 10 othersatlas.village.dev

Margins

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

AutomationBench-AA 1.0.6Tool & computer use › Customer service & operations36.068.1+32.1
HLAB-Vals-1.1Professional › Legal4.225.0+20.8
DAXBench v2Software › Data & ML81.962.8−19.1
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Professional › Legal4.922.7+17.8
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & medicine35.018.3−16.7
SimpleBenchKnowledge & reasoning › General knowledge59.272.7+13.5
KradleBenchGames15.025.7+10.7
τ³-Banking (Mercor run, bm25_grep retrieval)Tool & computer use › Customer service & operations43.053.3+10.3
Artificial Analysis Strategy & Ops IndexProfessional › General knowledge work39.148.7+9.6
Tax Agent BenchProfessional › Finance, accounting & tax62.271.4+9.1
BuseyBench SVGPerception & media › Generation4.65.4+8.4
Terminal-Bench 4.0 (Artificial Analysis)Software › Software engineering32.840.9+8.1
FrontierCode 2.0 (Artificial Analysis)Software › Software engineering34.927.2−7.8
APEX-Agents 1.1Professional › General knowledge work68.375.7+7.4
Vibe Code Bench v1.1Software › Frontend & design90.483.7−6.7
Excel Modeling Benchmark v1.1Professional › Finance, accounting & tax65.572.2+6.7
APEX-1 (Mercor, mean score)Professional › General knowledge work64.770.4+5.7
Ramp SWE-Bench v2Software › Software engineering57.451.8−5.6
Vals Legal IndexProfessional › Legal70.375.7+5.4
Legal Research BenchProfessional › Legal43.849.0+5.3
Humanity's Last Exam DiamondKnowledge & reasoning › General knowledge29.625.1−4.5
CritPt v2 (agentic, avg@5)Science › Physical sciences66.362.0−4.3
Artificial Analysis Healthcare & Medical IndexHealth & medicine37.541.3+3.8
DeepFinanceProfessional › Finance, accounting & tax40.036.5−3.5
Vals Finance IndexProfessional › Finance, accounting & tax57.460.9+3.5
SWE Atlas v1.1 - OverallSoftware › Software engineering59.456.0−3.4
Code Migration v1.1 (Vals Index v2.2 subset)Software › Software engineering50.247.1−3.1
LMArena Agent Arena Net Improvement (Steering Burden)Tool & computer use › General agents-0.61.0+3.1
Terminal-Bench 4.0Software › Software engineering36.439.4+3.0
AA-Briefcase v1.1 EloProfessional › General knowledge work15721520−2.8
FORTRESS Average Risk Score (ARS)Safety & alignment9.511.9−2.4
AA-Omniscience v1.1 IndexKnowledge & reasoning › General knowledge12.37.7−2.3
APEX-SWESoftware › Software engineering55.152.9−2.2
LMArena Agent Arena Steering BurdenTool & computer use › General agents0.84.8+2.0
GDP.pdf (Artificial Analysis, All-pass)Professional › General knowledge work20.822.8+2.0
TaxCalcBench TY25Professional › Finance, accounting & tax4.02.0−2.0
Vals Index v2.2Composite indexes58.760.5+1.7
Artificial Analysis Economics IndexComposite indexes42.741.0−1.7
LMArena Code Arena WebDevSoftware › Frontend & design15961608+1.7
Finance Agent v2.1Professional › Finance, accounting & tax65.463.8−1.5
LMArena Agent Arena Praise vs ComplaintTool & computer use › General agents-5.5-2.7+1.4
LMArena Agent Arena Tool HallucinationTool & computer use › Tool calling-1.50.9+1.2
LMArena Agent Arena Confirmed SuccessTool & computer use › General agents0.2-2.0−1.1
GBENCH One-shot Coding GScoreGames55.456.4+1.0
Artificial Analysis Finance & Accounting IndexProfessional › Finance, accounting & tax43.644.6+1.0
Artificial Analysis Progress Index v1Composite indexes326329+0.9
Artificial Analysis Engineering IndexScience › Engineering42.541.8−0.8
Artificial Analysis Intelligence Index v5.0.1Composite indexes36.536.0−0.5
Artificial Analysis Legal IndexProfessional › Legal38.738.5−0.2
LMArena Agent Arena Bash ReliabilityTool & computer use › General agents6.66.60.0
Terminal-Bench-Science 0.1Science › Multidisciplinary20.020.00.0
Data: Artificial Analysis, Vals AI, Arena and 10 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…