Model comparison
Nemotron 3 Ultra vs o4-mini
Nemotron 3 Ultra is the stronger model overall, scoring 44.7 to 41.6 on the Noometry Index.
Last verified . 25 shared benchmarks.
NVIDIASummary
- They share 25 benchmarks with published results for both. Nemotron 3 Ultra scores higher in 4 categories and o4-mini in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Nemotron 3 Ultra leads 66.3 to 54.2.
- The biggest single-benchmark swing is Mystery Game Puzzles: 20% for Nemotron 3 Ultra and 5% for o4-mini.
- Nemotron 3 Ultra is cheaper at $0.60 / $2.40 per million input/output tokens, against $1.10 / $4.40 for o4-mini.
- Nemotron 3 Ultra accepts more context: 262K tokens versus 200K.
- Nemotron 3 Ultra has downloadable open weights; the other is API-only.
Side by side
| Nemotron 3 Ultra | o4-mini | |
|---|---|---|
| Provider | NVIDIA | OpenAI |
| Noometry Index | 44.7 | 41.6 |
| Released | 2026-06-04 | 2025-04-16 |
| Weights | Open | Proprietary |
| Context window | 262K | 200K |
| Max output | 128K | 100K |
| Input $ / M tokens | $0.60 | $1.10 |
| Output $ / M tokens | $2.40 | $4.40 |
| Results tracked | 29 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding o4-mini leads
Nemotron 3 Ultra: 37.7 (#174), o4-mini: 41.6 (#110)
| Benchmark | Nemotron 3 Ultra | o4-mini |
|---|---|---|
| WeirdML | 43.5% | 52.6% |
| LMArena Coding | 1468 | 1369 |
| FrontierCode | 13.6% | — |
| SWE-bench Verified (bash only) | — | 45% |
| Aider Polyglot | — | 72% |
| SciCode | 40.3% | — |
| GSO | — | 3.6% |
| CadEval | — | 62% |
| ALE-Bench | — | 826.17 |
| AlgoTune | — | 1.72 |
Agentic & Tool Use Not comparable
Nemotron 3 Ultra: —, o4-mini: 32.3 (#57)
| Benchmark | Nemotron 3 Ultra | o4-mini |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 53.2% |
| GDPval | — | 25.3% |
| METR Time Horizons | — | 63.9% |
Reasoning Nemotron 3 Ultra leads
Nemotron 3 Ultra: 30.3 (#78), o4-mini: 24.6 (#143)
| Benchmark | Nemotron 3 Ultra | o4-mini |
|---|---|---|
| CritPt | 3.1% | 0.6% |
| Chess Puzzles | 12% | 26% |
| LMArena Hard Prompts | 1452 | 1351 |
| Mystery Game Puzzles | 20% | 5% |
| DTBench | 90.1% | 77.6% |
| LMCA | 36.9% | 26.5% |
| Epoch Capabilities Index | 146.2 | 145.63 |
| ARC-AGI-2 | — | 6.1% |
| SimpleBench | — | 38.7% |
| Kagi LLM Benchmark | — | 67.6% |
| ARC-AGI-1 | — | 58.7% |
| EnigmaEval | — | 9.2% |
| ForecastBench | — | 61.8 |
Math o4-mini leads
Nemotron 3 Ultra: 34.4 (#160), o4-mini: 40.2 (#79)
| Benchmark | Nemotron 3 Ultra | o4-mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 86.7% | 81.7% |
| LMArena Math | 1456 | 1390 |
| FrontierMath (Tiers 1-3) | — | 36.1% |
| FrontierMath Tier 4 | — | 4.9% |
| ProofBench | 2% | — |
| Omni-MATH | — | 72% |
| MATH Level 5 | — | 97.8% |
| FrontierMath (Feb 2025 set) | — | 24.8% |
| FrontierMath Tier 4 (v1) | — | 6.3% |
Knowledge Nemotron 3 Ultra leads
Nemotron 3 Ultra: 52.5 (#56), o4-mini: 43.5 (#83)
| Benchmark | Nemotron 3 Ultra | o4-mini |
|---|---|---|
| GPQA Diamond | 85.4% | 79.6% |
| LMArena Expert | 1472 | 1344 |
| Humanity's Last Exam | — | 18.1% |
| SimpleQA Verified | — | 19.6% |
| MMLU-Pro | — | 82% |
| Confabulations | — | 15.8% |
| Vectara Hallucination Rate | — | 18.6% |
| GPQA (HELM) | — | 73.5% |
Multimodal Not comparable
Nemotron 3 Ultra: —, o4-mini: 39.5 (#46)
| Benchmark | Nemotron 3 Ultra | o4-mini |
|---|---|---|
| LMArena Vision | — | 1194 |
| GeoBench | — | 64% |
| VPCT | — | 57.5% |
Multilingual Nemotron 3 Ultra leads
Nemotron 3 Ultra: 53.7 (#61), o4-mini: 47.2 (#149)
| Benchmark | Nemotron 3 Ultra | o4-mini |
|---|---|---|
| LMArena Non-English | 1427 | 1336 |
| LMArena Chinese | 1498 | 1355 |
| LMArena French | 1475 | 1362 |
| LMArena German | 1471 | 1336 |
| LMArena Korean | 1387 | 1313 |
| LMArena Russian | 1415 | 1334 |
| LMArena Spanish | 1452 | 1346 |
| LMArena Japanese | — | 1308 |
Instruction Following Too close to call
Nemotron 3 Ultra: 74.9 (#84), o4-mini: 75.4 (#63)
| Benchmark | Nemotron 3 Ultra | o4-mini |
|---|---|---|
| LMArena Instruction Following | 1417 | 1321 |
| IFEval | — | 92.8% |
Long Context o4-mini leads
Nemotron 3 Ultra: 43.9 (#77), o4-mini: 45.5 (#31)
| Benchmark | Nemotron 3 Ultra | o4-mini |
|---|---|---|
| LMArena Longer Query | 1437 | 1315 |
| Fiction.LiveBench | — | 77.8% |
Writing & Preference Nemotron 3 Ultra leads
Nemotron 3 Ultra: 66.3 (#35), o4-mini: 54.2 (#148)
| Benchmark | Nemotron 3 Ultra | o4-mini |
|---|---|---|
| LMArena Text | 1444 | 1353 |
| LMArena Creative Writing | 1392 | 1294 |
| LMArena Multi-Turn | 1412 | 1350 |
| Short-Story Creative Writing | — | 75% |
| EQ-Bench Creative Writing | 1692 | — |
| WildBench | — | 85.4% |
Frequently asked questions
Is Nemotron 3 Ultra better than o4-mini?
Nemotron 3 Ultra is the stronger model overall, scoring 44.7 to 41.6 on the Noometry Index.
Which is cheaper, Nemotron 3 Ultra or o4-mini?
Nemotron 3 Ultra is cheaper. It lists at $0.60 per million input tokens and $2.40 per million output tokens; o4-mini lists at $1.10 and $4.40.
Is Nemotron 3 Ultra or o4-mini better for coding?
o4-mini scores higher on coding benchmarks: 41.6 versus 37.7 in the Noometry coding category.
Which has the bigger context window?
Nemotron 3 Ultra does, with 262K tokens against 200K.
How many benchmarks do Nemotron 3 Ultra and o4-mini share?
25 benchmarks have published results for both models. Nemotron 3 Ultra has 29 scored results on Noometry and o4-mini has 60.