Atlas

Compare Models

MiniMax M3.1 Flash vs Muse Wonder

Muse Wonder leads by 3.7 capability points (≈95% interval +1.2 to +6.2): a clear lead over MiniMax M3.1 Flash · P(higher) ≈ 100%

Wins 33 of 35 shared benchmarks (2 lost)

Capability vs. relative cost
MiniMax M3.1 FlashMuse 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.

MiniMax M3.1 FlashvsMuse Wondercapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins2 – 33 · 35 shared

Comparing matching setups only (58 benchmarks excluded) · suites counted once
← MiniMax M3.1 FlashMuse Wonder →

MiniMax M3.1 FlashvsMuse Wonderwins by categoryData Nov 15, 2026Data: Vals AI, Mercor, Scale 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 Wonder minus MiniMax M3.1 Flash.

Visa Agentic Commerce BenchmarkAgents33.382.2+48.9
Toloka Arena (v3)Agents22.065.7+43.7
NoCoT-Bench NCKI (kspine v3)General QA65.9109+30.1
Terminal-Bench 4.0Code25.349.5+24.2
BuseyBench SVGMultimodal4.26.6+23.8
TaxCalcBench TY25Finance & Accounting0.022.0+22.0
Code Migration v1.1 (Vals Index v2.2 subset)Code48.269.8+21.6
DAXBench v2Code63.884.0+20.2
GBENCH One-shot Coding GScoreGames41.158.1+16.9
Legal Research BenchLegal34.150.5+16.3
HLAB-Vals-1.1Legal7.523.8+16.3
NoCoT-Bench NCRI 15.2General QA68.497.7+16.1
τ³-Banking (Mercor run, bm25_grep retrieval)Agents36.151.9+15.8
HiL-Bench (Human-in-Loop Benchmark)Agents65.779.7+14.0
Professional Reasoning Benchmark - LegalLegal45.158.4+13.3
Professional Reasoning Benchmark - FinanceFinance & Accounting45.658.7+13.0
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal11.923.0+11.1
Vals Index v2.2Indexes53.864.9+11.1
APEX-1 (Mercor, mean score)Professional Work65.075.1+10.1
Humanity's Last Exam DiamondGeneral QA28.137.3+9.2
Vibe Code Bench v1.1Code73.983.1+9.1
Agents' Last Exam v1.1 (ALE v1.1)Agents29.538.5+9.0
APEX-Agents 1.1Professional Work68.177.0+8.9
Tax Agent BenchFinance & Accounting62.771.2+8.4
Vals Legal IndexIndexes67.175.3+8.2
APEX-Accounting 1.1Finance & Accounting60.067.1+7.1
Finance Agent v2.1Finance & Accounting58.865.5+6.7
Vals Finance IndexIndexes55.961.2+5.3
DeepSWE v1.1Code73.778.8+5.1
LiveBench 2.0Agents73.178.1+5.0
Excel Modeling Benchmark v1.1Finance & Accounting67.872.5+4.7
FrontierML ($200 wallet)Agents23.219.3−4.0
GDPxProfessional Work4.06.8+2.8
ErdosBenchMath3.03.0+1.0
Access to Justice BenchmarkLegal73.373.0−0.4

MiniMax M3.1 FlashvsMuse Wondernormalised score marginsData Nov 15, 2026Data: Vals AI, Mercor, Scale 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…