Atlas

Compare Models

Claude Opus 5.5 vs MiniMax M3.1 Flash

Claude Opus 5.5 leads by 7.9 capability points (≈95% interval +5.2 to +10.5): a clear lead over MiniMax M3.1 Flash · P(higher) ≈ 100%

Wins 29 of 31 shared benchmarks (2 lost)

Capability vs. relative cost
Claude Opus 5.5MiniMax M3.1 FlashClosed 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 Opus 5.5vsMiniMax M3.1 Flashcapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins29 – 2 · 31 shared

Comparing matching setups only (45 benchmarks excluded) · suites counted once
← Claude Opus 5.5MiniMax M3.1 Flash →

Claude Opus 5.5vsMiniMax M3.1 Flashwins by categoryData Nov 15, 2026Data: Mercor, Vals AI, Scale AI and 19 othersVillage LLM Atlasatlas.village.dev

Margins

Each benchmark scaled 0–100% (its declared range, else the field’s range); Δ is MiniMax M3.1 Flash minus Claude Opus 5.5.

Toloka Arena (v3)Agents75.022.0−53.0
ALE-BenchCode27481495−44.4
TaxCalcBench TY25Finance & Accounting42.00.0−42.0
FrontierML ($200 wallet)Agents58.823.2−35.6
GBENCH One-shot Coding GScoreGames73.941.1−32.8
DAXBench v2Code95.763.8−31.9
Artificial Analysis Intelligence Index v5.0Indexes58.028.9−29.1
Visa Agentic Commerce BenchmarkAgents62.233.3−28.9
Hebbia Financial Services BenchmarkFinance & Accounting74.647.0−27.6
Humanity's Last Exam DiamondGeneral QA55.028.1−26.9
BuseyBench SVGMultimodal6.84.2−26.3
o11y-bench 1.1 Pass^3Agents84.464.1−20.3
τ³-Banking (Mercor run, bm25_grep retrieval)Agents55.036.1−18.9
Design Arena Overall Code CategoriesCode13841297−14.4
Agents' Last Exam v1.1 (ALE v1.1)Agents43.429.5−13.9
LiveBench 2.0Agents86.573.1−13.4
APEX-Agents 1.1Professional Work81.368.1−13.2
PostTrainBench Average (v1.1)Indexes49.339.3−10.0
Frontier3DBenchCode62.352.5−9.7
Agents' Last Exam (ALE)Agents38.229.6−8.6
GDPxProfessional Work12.04.0−8.0
Excel Modeling Benchmark v1.1Finance & Accounting75.867.8−8.0
HiL-Bench (Human-in-Loop Benchmark)Agents59.065.7+6.7
SAGE v1.1Math91.385.3−6.0
APEX-Accounting 1.1Finance & Accounting65.560.0−5.5
Professional Reasoning Benchmark - LegalLegal47.645.1−2.5
HLAB-Vals-1.1Legal9.27.5−1.7
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal10.811.9+1.1
APEX-1 (Mercor, mean score)Professional Work65.865.0−0.8
DeepSWE v1.1Code74.273.7−0.5
Access to Justice BenchmarkLegal73.873.3−0.5

Claude Opus 5.5vsMiniMax M3.1 Flashnormalised score marginsData Nov 15, 2026Data: Mercor, Vals AI, Scale AI and 19 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…