Atlas

Compare Models

Claude Fable 5.2 vs Muse Spark 1.3

Claude Fable 5.2 leads by 9.1 capability points (≈95% interval +6.4 to +11.7): a clear lead over Muse Spark 1.3 · P(higher) ≈ 100%

Wins 52 of 61 shared benchmarks (9 lost)

Capability vs. relative cost
Claude Fable 5.2Muse Spark 1.3Closed 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 Fable 5.2vsMuse Spark 1.3capability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins52 – 9 · 61 shared

Comparing matching setups only (7 benchmarks excluded) · suites counted once
← Claude Fable 5.2Muse Spark 1.3 →

Claude Fable 5.2vsMuse Spark 1.3wins by categoryData Nov 15, 2026Data: Artificial Analysis, Arena, Vals AI and 15 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is Muse Spark 1.3 minus Claude Fable 5.2.

HoneyBench v0.2 (hack rate)Safety11.983.6−71.7
AA-PersonalityBench Slop indexChat & Writing6.775.3+69.2
Mystery Game PuzzlesGames90.025.0−65.0
AA-PersonalityBench Informativeness indexChat & Writing46.7100+53.8
FrontierMath Tier 4 (v2) - PrivateMath100.046.3−53.7
LMArena Agent Arena Net Improvement (Steering Burden)Agents30.42.2−53.2
Humanity's Last Exam DiamondGeneral QA80.233.7−46.5
ChartographyMultimodal73.427.6−45.8
FrontierSWE V2Code69.925.8−44.1
ProofBench v1.1Math100.058.0−42.0
Chess Puzzles (task version 2.0)Games79.037.0−42.0
AA-PersonalityBench Transparency indexChat & Writing81.041.0−40.4
LMArena Agent Arena Praise vs ComplaintAgents73.41.7−35.9
AA-3D v1.0 EloCode12711032−34.8
AA-PersonalityBench Imagination indexChat & Writing67.698.1+30.8
GameBench 2Games99.168.5−30.6
Humanity's Last ExamGeneral QA73.243.2−30.0
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine43.915.6−28.3
APEX-Agents 1.1Professional Work87.661.3−26.3
FrontierCode 2.0 (Artificial Analysis)Code48.823.1−25.6
Bug Hunt BenchCode55.230.7−24.6
LMArena Code Arena WebDevCode18361655−23.7
AutomationBench 1.0.6 (Zapier)Agents44.320.7−23.6
FrontierMath Tiers 1-3 (v2) - PrivateMath95.174.4−20.7
SWE-bench Multimodal (Vals v1, 480 Tasks)Code62.343.1−19.2
Conceptual Reasoning IndexIndexes79.966.2−18.8
CritPt v2 (agentic, avg@5)Science86.368.0−18.3
Vals RSI Index v1.1 (four tasks)Agents37.919.6−18.2
LMCAGeneral QA70.653.9−16.7
AutomationBench-AA 1.0.6Agents74.257.9−16.3
MysteryMechanismScience51.836.0−15.8
FrontierCode 1.1 Main (Artificial Analysis)Code56.240.6−15.6
LMArena Agent Arena Steering BurdenAgents33.02.6−15.2
BuseyBench SVGMultimodal7.25.7−14.6
ACCoRDGeneral QA89.475.1−14.3
AA-PersonalityBench Warmth indexChat & Writing82.996.2+13.5
Epoch Capabilities Index (ECI)Indexes174158−13.2
MineBench — Bradley-Terry RatingGames19321747−11.2
AA-Briefcase v1.1 EloAgents17851581−11.2
CritPt (avg@5)Science35.726.0−9.7
Ramp Accounting BenchFinance & Accounting52.242.5−9.7
AA-PersonalityBench Organization indexChat & Writing54.363.8+9.6
AA-PersonalityBench Expressiveness indexChat & Writing81.071.4−9.6
LMArena Agent Arena Confirmed SuccessAgents22.03.7−9.1
Immigration BenchLegal74.065.0−9.0
SciCode 1.0.1 (Artificial Analysis)Code67.759.7−8.0
Vibe Code Bench v1.1Code93.885.9−7.9
GDPval-AA v2.1Professional Work18311681−6.6
DeepSWE v1.1Code74.367.8−6.5
Tax Agent BenchFinance & Accounting67.073.4+6.4
LMArena Agent Arena Bash ReliabilityAgents14.42.0−6.2
APEX-Accounting 1.1Finance & Accounting70.464.9−5.5
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal10.015.0+5.0
τ³-Banking (Mercor run, bm25_grep retrieval)Agents55.752.2−3.5
DAXBench v2Code88.385.1−3.2
DTBenchGeneral QA99.296.5−2.7
AA-PersonalityBench Conviction indexChat & Writing75.377.2+1.9
Harvey LAB-AA v1.1 (hallucination-gated)Legal7.28.9+1.7
OTIS Mock AIME 2024-2025Math99.799.2−0.6
AA-LCR v1.1Search83.383.0−0.3
LMArena Agent Arena Tool HallucinationAgents1.00.8−0.1

Claude Fable 5.2vsMuse Spark 1.3normalised score marginsData Nov 15, 2026Data: Artificial Analysis, Arena, Vals AI and 15 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…