Model comparison
Nemotron 3 Ultra vs Wizardlm 70b
Nemotron 3 Ultra is the stronger model overall, scoring 42.5 to 33.0 on the Noometry Index.
Last verified . 12 shared benchmarks.
NVIDIA
MicrosoftSummary
- They share 12 benchmarks with published results for both. Nemotron 3 Ultra scores higher in 7 categories and Wizardlm 70b in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Nemotron 3 Ultra leads 66.3 to 34.8.
Side by side
| Nemotron 3 Ultra | Wizardlm 70b | |
|---|---|---|
| Provider | NVIDIA | Microsoft |
| Noometry Index | 42.5 | 33.0 |
| Released | 2026-06-04 | — |
| Weights | Open | Open |
| Context window | 262K | — |
| Max output | 128K | — |
| Input $ / M tokens | $0.60 | — |
| Output $ / M tokens | $2.40 | — |
| Results tracked | 30 | 12 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Nemotron 3 Ultra leads
Nemotron 3 Ultra: 38.1 (#182), Wizardlm 70b: 31.4 (#285)
| Benchmark | Nemotron 3 Ultra | Wizardlm 70b |
|---|---|---|
| LMArena Coding | 1468 | 1081 |
| FrontierCode | 13.6% | — |
| SciCode | 40.3% | — |
| WeirdML | 43.5% | — |
Agentic & Tool Use Not comparable
Nemotron 3 Ultra: 23.4 (#126), Wizardlm 70b: —
| Benchmark | Nemotron 3 Ultra | Wizardlm 70b |
|---|---|---|
| APEX-Agents | 22.7% | — |
Reasoning Nemotron 3 Ultra leads
Nemotron 3 Ultra: 31.0 (#82), Wizardlm 70b: 20.7 (#234)
| Benchmark | Nemotron 3 Ultra | Wizardlm 70b |
|---|---|---|
| LMArena Hard Prompts | 1452 | 1079 |
| CritPt | 3.1% | — |
| Chess Puzzles | 12% | — |
| Mystery Game Puzzles | 20% | — |
| DTBench | 90.1% | — |
| LMCA | 36.9% | — |
| Epoch Capabilities Index | 146.17 | — |
Math Nemotron 3 Ultra leads
Nemotron 3 Ultra: 35.3 (#187), Wizardlm 70b: 32.2 (#218)
| Benchmark | Nemotron 3 Ultra | Wizardlm 70b |
|---|---|---|
| LMArena Math | 1457 | 1116 |
| OTIS Mock AIME 2024-2025 | 86.7% | — |
| ProofBench | 2% | — |
Knowledge Not comparable
Nemotron 3 Ultra: 52.5 (#61), Wizardlm 70b: —
| Benchmark | Nemotron 3 Ultra | Wizardlm 70b |
|---|---|---|
| GPQA Diamond | 85.4% | — |
| LMArena Expert | 1472 | — |
Multilingual Nemotron 3 Ultra leads
Nemotron 3 Ultra: 53.5 (#64), Wizardlm 70b: 29.6 (#265)
| Benchmark | Nemotron 3 Ultra | Wizardlm 70b |
|---|---|---|
| LMArena Non-English | 1427 | 1078 |
| LMArena Chinese | 1497 | 1052 |
| LMArena German | 1471 | 1083 |
| LMArena Russian | 1417 | 1155 |
| LMArena French | 1472 | — |
| LMArena Korean | 1386 | — |
| LMArena Spanish | 1454 | — |
Instruction Following Nemotron 3 Ultra leads
Nemotron 3 Ultra: 74.7 (#88), Wizardlm 70b: 56.3 (#273)
| Benchmark | Nemotron 3 Ultra | Wizardlm 70b |
|---|---|---|
| LMArena Instruction Following | 1418 | 1093 |
Long Context Nemotron 3 Ultra leads
Nemotron 3 Ultra: 43.9 (#81), Wizardlm 70b: 33.3 (#263)
| Benchmark | Nemotron 3 Ultra | Wizardlm 70b |
|---|---|---|
| LMArena Longer Query | 1437 | 1097 |
Writing & Preference Nemotron 3 Ultra leads
Nemotron 3 Ultra: 66.3 (#36), Wizardlm 70b: 34.8 (#269)
| Benchmark | Nemotron 3 Ultra | Wizardlm 70b |
|---|---|---|
| LMArena Text | 1445 | 1120 |
| LMArena Creative Writing | 1398 | 1149 |
| LMArena Multi-Turn | 1411 | 1108 |
| EQ-Bench Creative Writing | 1692 | — |
Frequently asked questions
Is Nemotron 3 Ultra better than Wizardlm 70b?
Nemotron 3 Ultra is the stronger model overall, scoring 42.5 to 33.0 on the Noometry Index.
Is Nemotron 3 Ultra or Wizardlm 70b better for coding?
Nemotron 3 Ultra scores higher on coding benchmarks: 38.1 versus 31.4 in the Noometry coding category.
How many benchmarks do Nemotron 3 Ultra and Wizardlm 70b share?
12 benchmarks have published results for both models. Nemotron 3 Ultra has 30 scored results on Noometry and Wizardlm 70b has 12.