Atlas

Compare Models

GPT-6.1 Sol vs MiniMax M3.1 Flash

GPT-6.1 Sol leads by 6.4 capability points (≈95% interval +3.8 to +9.0): a clear lead over MiniMax M3.1 Flash · P(higher) ≈ 100%

Wins 29 of 36 shared benchmarks (7 lost)

Capability vs. relative cost
GPT-6.1 SolMiniMax 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.

GPT-6.1 SolvsMiniMax M3.1 Flashcapability vs. relative costData Nov 15, 2026Village LLM Atlasatlas.village.dev

Benchmark wins29 – 7 · 36 shared

Comparing matching setups only (67 benchmarks excluded) · suites counted once
← GPT-6.1 SolMiniMax M3.1 Flash →

GPT-6.1 SolvsMiniMax M3.1 Flashwins 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 MiniMax M3.1 Flash minus GPT-6.1 Sol.

ALE-BenchCode29901495−53.0
NoCoT-Bench NCRI 15.2General QA14968.4−44.3
NoCoT-Bench NCKI (kspine v3)General QA13065.9−44.1
Toloka Arena (v3)Agents66.022.0−44.0
TaxCalcBench TY25Finance & Accounting38.00.0−38.0
GBENCH One-shot Coding GScoreGames79.041.1−37.9
BuseyBench SVGMultimodal8.04.2−37.3
Terminal-Bench 4.0Code55.125.3−29.8
Visa Agentic Commerce BenchmarkAgents62.233.3−28.9
DAXBench v2Code92.663.8−28.7
Code Migration v1.1 (Vals Index v2.2 subset)Code75.148.2−26.9
Hebbia Financial Services BenchmarkFinance & Accounting73.547.0−26.5
o11y-bench 1.1 Pass^3Agents85.964.1−21.9
Vibe Code Bench v1.1Code88.973.9−15.0
LiveBench 2.0Agents85.473.1−12.3
Vals Index v2.2Indexes63.953.8−10.2
Design Arena Overall Code CategoriesCode13561297−9.8
Access to Justice BenchmarkLegal65.873.3+7.5
τ³-Banking (Mercor run, bm25_grep retrieval)Agents43.036.1−6.9
SAGE v1.1Math90.585.3−5.2
APEX-Agents 1.1Professional Work73.268.1−5.1
GDPxProfessional Work8.44.0−4.4
Legal Research BenchLegal38.534.1−4.3
APEX-Accounting 1.1Finance & Accounting64.360.0−4.3
Agents' Last Exam Overall ScoreAgents58.655.0−3.6
HiL-Bench (Human-in-Loop Benchmark)Agents62.365.7+3.3
Excel Modeling Benchmark v1.1Finance & Accounting71.167.8−3.2
Harvey LAB (Mercor run, undisclosed ~213-task cohort)Legal8.811.9+3.1
Finance Agent v2.1Finance & Accounting61.858.8−3.0
Professional Reasoning Benchmark - LegalLegal47.945.1−2.8
Vals Legal IndexIndexes68.767.1−1.6
HLAB-Vals-1.1Legal6.37.5+1.3
Professional Reasoning Benchmark - FinanceFinance & Accounting46.245.6−0.6
SWE-Bench Pro V2 — FullCode96.196.6+0.5
Tax Agent BenchFinance & Accounting62.362.7+0.4
Vals Finance IndexIndexes55.855.9+0.1

GPT-6.1 SolvsMiniMax M3.1 Flashnormalised 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…