Model comparison
GPT-4o vs Nemotron 3 Ultra
Nemotron 3 Ultra is the stronger model overall, scoring 44.7 to 28.5 on the Noometry Index.
Last verified . 24 shared benchmarks.
NVIDIASummary
- They share 24 benchmarks with published results for both. GPT-4o scores higher in 0 categories and Nemotron 3 Ultra in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Nemotron 3 Ultra leads 34.4 to 10.2.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 6.4% for GPT-4o and 86.7% for Nemotron 3 Ultra.
- Nemotron 3 Ultra is cheaper at $0.60 / $2.40 per million input/output tokens, against $2.50 / $10 for GPT-4o.
- Nemotron 3 Ultra accepts more context: 262K tokens versus 128K.
- Nemotron 3 Ultra has downloadable open weights; the other is API-only.
Side by side
| GPT-4o | Nemotron 3 Ultra | |
|---|---|---|
| Provider | OpenAI | NVIDIA |
| Noometry Index | 28.5 | 44.7 |
| Released | 2024-05-13 | 2026-06-04 |
| Weights | Proprietary | Open |
| Context window | 128K | 262K |
| Max output | 16K | 128K |
| Input $ / M tokens | $2.50 | $0.60 |
| Output $ / M tokens | $10 | $2.40 |
| Results tracked | 72 | 29 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Nemotron 3 Ultra leads
GPT-4o: 25.1 (#323), Nemotron 3 Ultra: 37.7 (#174)
| Benchmark | GPT-4o | Nemotron 3 Ultra |
|---|---|---|
| WeirdML | 25.1% | 43.5% |
| LMArena Coding | 1297 | 1468 |
| SWE-bench Verified | 31% | — |
| FrontierCode | — | 13.6% |
| SWE-bench Verified (bash only) | 21.6% | — |
| Aider Polyglot | 45.3% | — |
| SciCode | — | 40.3% |
| GSO | 0% | — |
| BigCodeBench Instruct | 51.1% | — |
| LiveBench Coding | 51.4% | — |
| BigCodeBench Complete | 61.1% | — |
| CadEval | 26% | — |
| HumanEval+ | 87.2% | — |
| MBPP+ | 72.2% | — |
Agentic & Tool Use Not comparable
GPT-4o: 20.8 (#133), Nemotron 3 Ultra: —
| Benchmark | GPT-4o | Nemotron 3 Ultra |
|---|---|---|
| GDPval | 9.9% | — |
| TheAgentCompany | 8.6% | — |
| Cybench | 12.5% | — |
| BALROG | 32.3% | — |
| LMArena Search | 1006 | — |
| METR Time Horizons | 40.8% | — |
Reasoning Nemotron 3 Ultra leads
GPT-4o: 9.4 (#338), Nemotron 3 Ultra: 30.3 (#78)
| Benchmark | GPT-4o | Nemotron 3 Ultra |
|---|---|---|
| CritPt | 0% | 3.1% |
| Chess Puzzles | 13% | 12% |
| LMArena Hard Prompts | 1281 | 1452 |
| DTBench | 64.5% | 90.1% |
| LMCA | 16.6% | 36.9% |
| Epoch Capabilities Index | 128.97 | 146.2 |
| ARC-AGI-2 | 0% | — |
| SimpleBench | 17.8% | — |
| ARC-AGI-1 | 4.5% | — |
| EnigmaEval | 0.8% | — |
| LiveBench Reasoning | 55.8% | — |
| Mystery Game Puzzles | — | 20% |
| LiveBench Data Analysis | 60.9% | — |
| ForecastBench | 57.7 | — |
| LiveBench | 55.3% | — |
Math Nemotron 3 Ultra leads
GPT-4o: 10.2 (#303), Nemotron 3 Ultra: 34.4 (#160)
| Benchmark | GPT-4o | Nemotron 3 Ultra |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 6.4% | 86.7% |
| LMArena Math | 1285 | 1456 |
| FrontierMath (Tiers 1-3) | 0.4% | — |
| ProofBench | — | 2% |
| Omni-MATH | 29.3% | — |
| LiveBench Math | 49.5% | — |
| MATH Level 5 | 53.3% | — |
| FrontierMath (Feb 2025 set) | 0.3% | — |
Knowledge Nemotron 3 Ultra leads
GPT-4o: 28.8 (#235), Nemotron 3 Ultra: 52.5 (#56)
| Benchmark | GPT-4o | Nemotron 3 Ultra |
|---|---|---|
| GPQA Diamond | 49.2% | 85.4% |
| LMArena Expert | 1250 | 1472 |
| Humanity's Last Exam | 2.7% | — |
| SimpleQA Verified | 26% | — |
| MMLU-Pro | 71.3% | — |
| Confabulations | 15.3% | — |
| Vectara Hallucination Rate | 9.6% | — |
| GPQA (HELM) | 52% | — |
| MMLU | 88.1% | — |
Multimodal Not comparable
GPT-4o: 33.9 (#85), Nemotron 3 Ultra: —
| Benchmark | GPT-4o | Nemotron 3 Ultra |
|---|---|---|
| LMArena Vision | 1137 | — |
| Video-MME | 71.9% | — |
| GeoBench | 71% | — |
| VPCT | 40% | — |
| ScienceQA | 88.5% | — |
Multilingual Nemotron 3 Ultra leads
GPT-4o: 43.4 (#181), Nemotron 3 Ultra: 53.7 (#61)
| Benchmark | GPT-4o | Nemotron 3 Ultra |
|---|---|---|
| LMArena Non-English | 1283 | 1427 |
| LMArena Chinese | 1277 | 1498 |
| LMArena French | 1303 | 1475 |
| LMArena German | 1282 | 1471 |
| LMArena Korean | 1234 | 1387 |
| LMArena Russian | 1286 | 1415 |
| LMArena Spanish | 1292 | 1452 |
| LMArena Japanese | 1257 | — |
Instruction Following Nemotron 3 Ultra leads
GPT-4o: 66.8 (#202), Nemotron 3 Ultra: 74.9 (#84)
| Benchmark | GPT-4o | Nemotron 3 Ultra |
|---|---|---|
| LMArena Instruction Following | 1278 | 1417 |
| LiveBench Instruction Following | 68.6% | — |
| IFEval | 81.7% | — |
Long Context Nemotron 3 Ultra leads
GPT-4o: 39.4 (#174), Nemotron 3 Ultra: 43.9 (#77)
| Benchmark | GPT-4o | Nemotron 3 Ultra |
|---|---|---|
| LMArena Longer Query | 1289 | 1437 |
| Fiction.LiveBench | 66.7% | — |
Writing & Preference Nemotron 3 Ultra leads
GPT-4o: 52.8 (#160), Nemotron 3 Ultra: 66.3 (#35)
| Benchmark | GPT-4o | Nemotron 3 Ultra |
|---|---|---|
| LMArena Text | 1300 | 1444 |
| LMArena Creative Writing | 1291 | 1392 |
| LMArena Multi-Turn | 1302 | 1412 |
| Short-Story Creative Writing | 81.8% | — |
| EQ-Bench Creative Writing | — | 1692 |
| WildBench | 82.8% | — |
| LiveBench Language | 47.6% | — |
Frequently asked questions
Is GPT-4o better than Nemotron 3 Ultra?
Nemotron 3 Ultra is the stronger model overall, scoring 44.7 to 28.5 on the Noometry Index.
Which is cheaper, GPT-4o or Nemotron 3 Ultra?
Nemotron 3 Ultra is cheaper. It lists at $0.60 per million input tokens and $2.40 per million output tokens; GPT-4o lists at $2.50 and $10.
Is GPT-4o or Nemotron 3 Ultra better for coding?
Nemotron 3 Ultra scores higher on coding benchmarks: 37.7 versus 25.1 in the Noometry coding category.
Which has the bigger context window?
Nemotron 3 Ultra does, with 262K tokens against 128K.
How many benchmarks do GPT-4o and Nemotron 3 Ultra share?
24 benchmarks have published results for both models. GPT-4o has 72 scored results on Noometry and Nemotron 3 Ultra has 29.