Atlas

Compare Models

Claude Fable 5.2 vs Muse Wonder

Claude Fable 5.2 leads by 6.7 capability points (≈95% interval +3.9 to +9.4): a clear lead over Muse Wonder · P(higher) ≈ 100%

Wins 58 of 69 shared benchmarks (11 lost)

Capability vs. relative cost
Claude Fable 5.2Muse 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.

Claude Fable 5.2vsMuse Wondercapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins58 – 11 · 69 shared

Comparing matching setups only (4 benchmarks excluded) · suites counted once
← Claude Fable 5.2Muse Wonder →

Claude Fable 5.2vsMuse Wonderwins by categoryData Nov 15, 2026Data: Artificial Analysis, Vals AI, Epoch AI and 18 othersVillage LLM Atlasatlas.village.dev

Margins

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

HoneyBench v0.2 (hack rate)Safety11.981.0−69.1
EBR-bench v4 (card ban, single agent)Games90.531.4−59.0
LMArena Agent Arena Net Improvement (Steering Burden)Agents30.41.8−54.0
AA-PersonalityBench Informativeness indexChat & Writing46.794.3+48.1
AA-PersonalityBench Slop indexChat & Writing6.754.3+48.1
OpenWorldEngineer (Mean@8)Code54.612.7−41.8
Humanity's Last Exam DiamondGeneral QA80.239.0−41.2
Mystery Game PuzzlesGames90.051.0−39.0
LMArena Agent Arena Praise vs ComplaintAgents73.4-1.8−37.6
AA-PersonalityBench Organization indexChat & Writing54.390.5+36.5
Furniture Assembly BenchmarkMultimodal90.054.2−35.8
FrontierMath Tier 4 (v2) - PrivateMath100.070.7−29.3
FrontierCode 2.0 - Pass RateCode54.728.8−25.9
Humanity's Last ExamGeneral QA73.247.6−25.6
HardGeoBench Avg DistanceMultimodal5521369−25.2
Bug Hunt BenchCode55.230.2−25.1
AA-PersonalityBench Transparency indexChat & Writing81.056.2−25.0
FrontierCode 2.0 - ScoreCode50.225.3−24.9
LMArena Code Arena WebDevCode18361649−24.5
MysteryMechanismScience51.827.5−24.3
AA-PersonalityBench Imagination indexChat & Writing67.690.5+23.1
Medical Long Context Reasoning (MLCR-AA, October 2026 methodology)Health & Medicine43.922.2−21.7
GameBench 2Games99.177.8−21.3
HardGeoBench <=100miMultimodal60.038.8−21.2
HardGeoBench <=10miMultimodal48.827.5−21.2
HLAB-Vals-1.1Legal5.023.8+18.8
HardGeoBench Country AccuracyMultimodal70.052.5−17.5
AA-PersonalityBench Warmth indexChat & Writing82.9100+17.3
Conceptual Reasoning IndexIndexes79.967.7−16.7
SWE-bench Multimodal (Vals v1, 480 Tasks)Code62.345.6−16.7
BioMysteryBench (Vals Terminus 2)Science83.767.0−16.7
ProofBench v1.1Math100.084.0−16.0
LMArena Agent Arena Steering BurdenAgents33.02.6−15.2
CritPt v2 (agentic, avg@5)Science86.371.1−15.1
CrossSearch (pass^3)Search56.042.0−14.0
LMCAGeneral QA70.656.8−13.8
ACCoRDGeneral QA89.476.2−13.2
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal10.023.0+13.0
FrontierMath Tiers 1-3 (v2) - PrivateMath95.182.1−13.0
Vibe Code Bench v1.1Code93.883.1−10.7
APEX-Agents 1.1Professional Work87.677.0−10.6
Epoch Capabilities Index (ECI)Indexes174162−9.9
FrontierCode 1.1 Main (Artificial Analysis)Code56.246.5−9.7
GDPxProfessional Work16.46.8−9.6
LMArena Agent Arena Confirmed SuccessAgents22.05.4−8.3
AA-PersonalityBench Expressiveness indexChat & Writing81.088.6+7.7
CritPt (avg@5)Science35.728.3−7.4
SciCode 1.0.1 (Artificial Analysis)Code67.760.3−7.4
AA-Briefcase v1.1 EloAgents17851659−6.9
MacroscopeBenchCode71.164.2−6.9
AutomationBench-AA 1.0.6Agents74.268.1−6.1
BuseyBench SVGMultimodal7.26.6−5.7
Excel Modeling Benchmark v1.1Finance & Accounting77.472.5−4.9
LMArena Agent Arena Bash ReliabilityAgents14.44.8−4.8
Immigration BenchLegal74.069.6−4.4
DAXBench v2Code88.384.0−4.3
Tax Agent BenchFinance & Accounting67.071.2+4.2
DTBenchGeneral QA99.295.2−4.0
Finance Agent v2.1Finance & Accounting69.565.5−3.9
MineBench — Bradley-Terry RatingGames19321868−3.9
AA-PersonalityBench Conviction indexChat & Writing75.379.1+3.8
τ³-Banking (Mercor run, bm25_grep retrieval)Agents55.751.9−3.8
Ramp Accounting BenchFinance & Accounting52.248.7−3.5
APEX-Accounting 1.1Finance & Accounting70.467.1−3.3
GDPval-AA v2.1Professional Work18311770−2.7
Vals Finance IndexIndexes63.661.2−2.4
DualEntry Accounting AI BenchmarkFinance & Accounting81.283.2+2.0
LMArena Agent Arena Tool HallucinationAgents1.0-0.5−0.8
OTIS Mock AIME 2024-2025Math99.799.4−0.3

Claude Fable 5.2vsMuse Wondernormalised score marginsData Nov 15, 2026Data: Artificial Analysis, Vals AI, Epoch AI and 18 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…