Model comparison
MiniMax-M3 vs Trinity Large Thinking
MiniMax-M3 is the stronger model overall, scoring 43.5 to 38.1 on the Noometry Index.
Last verified . 22 shared benchmarks.
MiniMax
Arcee AISummary
- They share 22 benchmarks with published results for both. MiniMax-M3 scores higher in 8 categories and Trinity Large Thinking in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where MiniMax-M3 leads 58.4 to 40.3.
- The biggest single-benchmark swing is NYT Connections (extended): 65.1% for MiniMax-M3 and 16.5% for Trinity Large Thinking.
- Trinity Large Thinking is cheaper at $0.25 / $0.80 per million input/output tokens, against $0.30 / $1.20 for MiniMax-M3.
- MiniMax-M3 accepts more context: 1M tokens versus 262K.
Side by side
| MiniMax-M3 | Trinity Large Thinking | |
|---|---|---|
| Provider | MiniMax | Arcee AI |
| Noometry Index | 43.5 | 38.1 |
| Released | 2026-06-01 | 2026-04-01 |
| Weights | Open | Open |
| Context window | 1M | 262K |
| Max output | 512K | 80K |
| Input $ / M tokens | $0.30 | $0.25 |
| Output $ / M tokens | $1.20 | $0.80 |
| Results tracked | 41 | 24 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M3 leads
MiniMax-M3: 41.5 (#111), Trinity Large Thinking: 33.8 (#238)
| Benchmark | MiniMax-M3 | Trinity Large Thinking |
|---|---|---|
| LMArena WebDev | 1482 | 1238 |
| SciCode | 47.1% | 36.1% |
| LMArena Coding | 1470 | 1381 |
| FrontierCode | 14.7% | — |
| ALE-Bench | 640.02 | — |
Agentic & Tool Use Not comparable
MiniMax-M3: 22.5 (#121), Trinity Large Thinking: —
| Benchmark | MiniMax-M3 | Trinity Large Thinking |
|---|---|---|
| APEX-Agents | 37.7% | — |
| OSWorld 2.0 | 4.6% | — |
| GBAEval | 0.9% | — |
| Vending-Bench 2 | 2,158 | — |
Reasoning MiniMax-M3 leads
MiniMax-M3: 29.3 (#83), Trinity Large Thinking: 16.3 (#298)
| Benchmark | MiniMax-M3 | Trinity Large Thinking |
|---|---|---|
| NYT Connections (extended) | 65.1% | 16.5% |
| CritPt | 3.7% | 0.9% |
| LMArena Hard Prompts | 1447 | 1350 |
| Surface Evolver Bench | 55% | 15.6% |
| SimpleBench | 45.8% | — |
| Chess Puzzles | 14% | — |
| Thematic Generalization | — | 41.6% |
| Mystery Game Puzzles | 8% | — |
| DTBench | 78.9% | — |
| LMCA | 33.7% | — |
| Epoch Capabilities Index | 147 | — |
| ForecastBench | 61.4 | — |
Math MiniMax-M3 leads
MiniMax-M3: 39.1 (#83), Trinity Large Thinking: 34.9 (#149)
| Benchmark | MiniMax-M3 | Trinity Large Thinking |
|---|---|---|
| LMArena Math | 1432 | 1366 |
| OTIS Mock AIME 2024-2025 | 71.1% | — |
| ProofBench | 18% | — |
Knowledge MiniMax-M3 leads
MiniMax-M3: 58.4 (#32), Trinity Large Thinking: 40.3 (#107)
| Benchmark | MiniMax-M3 | Trinity Large Thinking |
|---|---|---|
| LMArena Expert | 1462 | 1359 |
| GPQA Diamond | 90.9% | — |
| Vectara Hallucination Rate | — | 6.9% |
Multimodal Not comparable
MiniMax-M3: 39.5 (#45), Trinity Large Thinking: —
| Benchmark | MiniMax-M3 | Trinity Large Thinking |
|---|---|---|
| LMArena Vision | 1254 | — |
| LMArena Document | 1435 | — |
Multilingual MiniMax-M3 leads
MiniMax-M3: 53.2 (#70), Trinity Large Thinking: 46.4 (#155)
| Benchmark | MiniMax-M3 | Trinity Large Thinking |
|---|---|---|
| LMArena Non-English | 1420 | 1325 |
| LMArena Chinese | 1465 | 1372 |
| LMArena French | 1445 | 1374 |
| LMArena German | 1424 | 1356 |
| LMArena Japanese | 1381 | 1312 |
| LMArena Korean | 1372 | 1304 |
| LMArena Russian | 1430 | 1338 |
| LMArena Spanish | 1431 | 1356 |
Instruction Following MiniMax-M3 leads
MiniMax-M3: 75.7 (#56), Trinity Large Thinking: 70.7 (#157)
| Benchmark | MiniMax-M3 | Trinity Large Thinking |
|---|---|---|
| LMArena Instruction Following | 1433 | 1334 |
Long Context MiniMax-M3 leads
MiniMax-M3: 44.2 (#68), Trinity Large Thinking: 41.2 (#139)
| Benchmark | MiniMax-M3 | Trinity Large Thinking |
|---|---|---|
| LMArena Longer Query | 1445 | 1355 |
Writing & Preference MiniMax-M3 leads
MiniMax-M3: 62.3 (#79), Trinity Large Thinking: 54.0 (#152)
| Benchmark | MiniMax-M3 | Trinity Large Thinking |
|---|---|---|
| LMArena Text | 1433 | 1340 |
| LMArena Creative Writing | 1405 | 1321 |
| LMArena Multi-Turn | 1441 | 1342 |
| EQ-Bench 4 | 1150 | — |
Frequently asked questions
Is MiniMax-M3 better than Trinity Large Thinking?
MiniMax-M3 is the stronger model overall, scoring 43.5 to 38.1 on the Noometry Index.
Which is cheaper, MiniMax-M3 or Trinity Large Thinking?
Trinity Large Thinking is cheaper. It lists at $0.25 per million input tokens and $0.80 per million output tokens; MiniMax-M3 lists at $0.30 and $1.20.
Is MiniMax-M3 or Trinity Large Thinking better for coding?
MiniMax-M3 scores higher on coding benchmarks: 41.5 versus 33.8 in the Noometry coding category.
Which has the bigger context window?
MiniMax-M3 does, with 1M tokens against 262K.
How many benchmarks do MiniMax-M3 and Trinity Large Thinking share?
22 benchmarks have published results for both models. MiniMax-M3 has 41 scored results on Noometry and Trinity Large Thinking has 24.