Model comparison
MiniMax-M3 vs Nemotron 3 Ultra
Nemotron 3 Ultra is the stronger model overall, scoring 44.7 to 43.5 on the Noometry Index. MiniMax-M3 costs 2.0× less per token, which makes it the better buy when Nemotron 3 Ultra's lead doesn't matter for your workload.
Last verified . 27 shared benchmarks.
MiniMax
NVIDIASummary
- They share 27 benchmarks with published results for both. MiniMax-M3 scores higher in 5 categories and Nemotron 3 Ultra in 3 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where MiniMax-M3 leads 58.4 to 52.5.
- The biggest single-benchmark swing is ProofBench: 18% for MiniMax-M3 and 2% for Nemotron 3 Ultra.
- MiniMax-M3 is cheaper at $0.30 / $1.20 per million input/output tokens, against $0.60 / $2.40 for Nemotron 3 Ultra.
- MiniMax-M3 accepts more context: 1M tokens versus 262K.
Side by side
| MiniMax-M3 | Nemotron 3 Ultra | |
|---|---|---|
| Provider | MiniMax | NVIDIA |
| Noometry Index | 43.5 | 44.7 |
| Released | 2026-06-01 | 2026-06-04 |
| Weights | Open | Open |
| Context window | 1M | 262K |
| Max output | 512K | 128K |
| Input $ / M tokens | $0.30 | $0.60 |
| Output $ / M tokens | $1.20 | $2.40 |
| Results tracked | 41 | 29 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M3 leads
MiniMax-M3: 41.5 (#111), Nemotron 3 Ultra: 37.7 (#174)
| Benchmark | MiniMax-M3 | Nemotron 3 Ultra |
|---|---|---|
| FrontierCode | 14.7% | 13.6% |
| SciCode | 47.1% | 40.3% |
| LMArena Coding | 1470 | 1468 |
| LMArena WebDev | 1482 | — |
| WeirdML | — | 43.5% |
| ALE-Bench | 640.02 | — |
Agentic & Tool Use Not comparable
MiniMax-M3: 22.5 (#121), Nemotron 3 Ultra: —
| Benchmark | MiniMax-M3 | Nemotron 3 Ultra |
|---|---|---|
| APEX-Agents | 37.7% | — |
| OSWorld 2.0 | 4.6% | — |
| GBAEval | 0.9% | — |
| Vending-Bench 2 | 2,158 | — |
Reasoning Too close to call
MiniMax-M3: 29.3 (#83), Nemotron 3 Ultra: 30.3 (#78)
| Benchmark | MiniMax-M3 | Nemotron 3 Ultra |
|---|---|---|
| CritPt | 3.7% | 3.1% |
| Chess Puzzles | 14% | 12% |
| LMArena Hard Prompts | 1447 | 1452 |
| Mystery Game Puzzles | 8% | 20% |
| DTBench | 78.9% | 90.1% |
| LMCA | 33.7% | 36.9% |
| Epoch Capabilities Index | 147 | 146.2 |
| SimpleBench | 45.8% | — |
| NYT Connections (extended) | 65.1% | — |
| Surface Evolver Bench | 55% | — |
| ForecastBench | 61.4 | — |
Math MiniMax-M3 leads
MiniMax-M3: 39.1 (#83), Nemotron 3 Ultra: 34.4 (#160)
| Benchmark | MiniMax-M3 | Nemotron 3 Ultra |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 71.1% | 86.7% |
| ProofBench | 18% | 2% |
| LMArena Math | 1432 | 1456 |
Knowledge MiniMax-M3 leads
MiniMax-M3: 58.4 (#32), Nemotron 3 Ultra: 52.5 (#56)
| Benchmark | MiniMax-M3 | Nemotron 3 Ultra |
|---|---|---|
| GPQA Diamond | 90.9% | 85.4% |
| LMArena Expert | 1462 | 1472 |
Multimodal Not comparable
MiniMax-M3: 39.5 (#45), Nemotron 3 Ultra: —
| Benchmark | MiniMax-M3 | Nemotron 3 Ultra |
|---|---|---|
| LMArena Vision | 1254 | — |
| LMArena Document | 1435 | — |
Multilingual Too close to call
MiniMax-M3: 53.2 (#70), Nemotron 3 Ultra: 53.7 (#61)
| Benchmark | MiniMax-M3 | Nemotron 3 Ultra |
|---|---|---|
| LMArena Non-English | 1420 | 1427 |
| LMArena Chinese | 1465 | 1498 |
| LMArena French | 1445 | 1475 |
| LMArena German | 1424 | 1471 |
| LMArena Korean | 1372 | 1387 |
| LMArena Russian | 1430 | 1415 |
| LMArena Spanish | 1431 | 1452 |
| LMArena Japanese | 1381 | — |
Instruction Following Too close to call
MiniMax-M3: 75.7 (#56), Nemotron 3 Ultra: 74.9 (#84)
| Benchmark | MiniMax-M3 | Nemotron 3 Ultra |
|---|---|---|
| LMArena Instruction Following | 1433 | 1417 |
Long Context Too close to call
MiniMax-M3: 44.2 (#68), Nemotron 3 Ultra: 43.9 (#77)
| Benchmark | MiniMax-M3 | Nemotron 3 Ultra |
|---|---|---|
| LMArena Longer Query | 1445 | 1437 |
Writing & Preference Nemotron 3 Ultra leads
MiniMax-M3: 62.3 (#79), Nemotron 3 Ultra: 66.3 (#35)
| Benchmark | MiniMax-M3 | Nemotron 3 Ultra |
|---|---|---|
| LMArena Text | 1433 | 1444 |
| LMArena Creative Writing | 1405 | 1392 |
| LMArena Multi-Turn | 1441 | 1412 |
| EQ-Bench Creative Writing | — | 1692 |
| EQ-Bench 4 | 1150 | — |
Frequently asked questions
Is MiniMax-M3 better than Nemotron 3 Ultra?
Nemotron 3 Ultra is the stronger model overall, scoring 44.7 to 43.5 on the Noometry Index. MiniMax-M3 costs 2.0× less per token, which makes it the better buy when Nemotron 3 Ultra's lead doesn't matter for your workload.
Which is cheaper, MiniMax-M3 or Nemotron 3 Ultra?
MiniMax-M3 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; Nemotron 3 Ultra lists at $0.60 and $2.40.
Is MiniMax-M3 or Nemotron 3 Ultra better for coding?
MiniMax-M3 scores higher on coding benchmarks: 41.5 versus 37.7 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 Nemotron 3 Ultra share?
27 benchmarks have published results for both models. MiniMax-M3 has 41 scored results on Noometry and Nemotron 3 Ultra has 29.