Model comparison
Nemotron 3 Nano 30B A3B vs Phi-4
Nemotron 3 Nano 30B A3B is the stronger model overall, scoring 40.6 to 31.2 on the Noometry Index.
Last verified . 18 shared benchmarks.
NVIDIA
MicrosoftSummary
- They share 18 benchmarks with published results for both. Nemotron 3 Nano 30B A3B scores higher in 8 categories and Phi-4 in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Nemotron 3 Nano 30B A3B leads 37.8 to 20.8.
- The biggest single-benchmark swing is Vectara Hallucination Rate: 9.6% for Nemotron 3 Nano 30B A3B and 3.7% for Phi-4.
- Both cost about the same: $0.05 input and $0.20 output per million tokens.
- Nemotron 3 Nano 30B A3B accepts more context: 262K tokens versus 128K.
Side by side
| Nemotron 3 Nano 30B A3B | Phi-4 | |
|---|---|---|
| Provider | NVIDIA | Microsoft |
| Noometry Index | 40.6 | 31.2 |
| Released | 2025-12-15 | 2024-12-11 |
| Weights | Open | Open |
| Context window | 262K | 128K |
| Max output | 262K | 4K |
| Input $ / M tokens | $0.05 | $0.07 |
| Output $ / M tokens | $0.20 | $0.14 |
| Results tracked | 18 | 37 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Nemotron 3 Nano 30B A3B leads
Nemotron 3 Nano 30B A3B: 40.5 (#138), Phi-4: 34.4 (#239)
| Benchmark | Nemotron 3 Nano 30B A3B | Phi-4 |
|---|---|---|
| LMArena Coding | 1378 | 1231 |
| BigCodeBench Instruct | — | 45.5% |
| LiveBench Coding | — | 30.7% |
| BigCodeBench Complete | — | 55.4% |
Agentic & Tool Use Not comparable
Nemotron 3 Nano 30B A3B: —, Phi-4: 22.8 (#128)
| Benchmark | Nemotron 3 Nano 30B A3B | Phi-4 |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 28.8% |
| BALROG | — | 11.6% |
Reasoning Nemotron 3 Nano 30B A3B leads
Nemotron 3 Nano 30B A3B: 27.1 (#124), Phi-4: 17.7 (#291)
| Benchmark | Nemotron 3 Nano 30B A3B | Phi-4 |
|---|---|---|
| LMArena Hard Prompts | 1348 | 1220 |
| Chess Puzzles | — | 1% |
| LiveBench Reasoning | — | 47.8% |
| LiveBench Data Analysis | — | 45.2% |
| Epoch Capabilities Index | — | 130.42 |
| LiveBench | — | 41.6% |
Math Nemotron 3 Nano 30B A3B leads
Nemotron 3 Nano 30B A3B: 37.8 (#147), Phi-4: 20.8 (#285)
| Benchmark | Nemotron 3 Nano 30B A3B | Phi-4 |
|---|---|---|
| LMArena Math | 1371 | 1246 |
| OTIS Mock AIME 2024-2025 | — | 13.8% |
| LiveBench Math | — | 42% |
| MATH Level 5 | — | 64.9% |
Knowledge Nemotron 3 Nano 30B A3B leads
Nemotron 3 Nano 30B A3B: 38.2 (#148), Phi-4: 32.6 (#209)
| Benchmark | Nemotron 3 Nano 30B A3B | Phi-4 |
|---|---|---|
| Vectara Hallucination Rate | 9.6% | 3.7% |
| LMArena Expert | 1355 | 1203 |
| GPQA Diamond | — | 56.1% |
| Confabulations | — | 29.4% |
| MMLU | — | 84.8% |
Multilingual Nemotron 3 Nano 30B A3B leads
Nemotron 3 Nano 30B A3B: 45.8 (#162), Phi-4: 37.2 (#237)
| Benchmark | Nemotron 3 Nano 30B A3B | Phi-4 |
|---|---|---|
| LMArena Non-English | 1320 | 1197 |
| LMArena Chinese | 1382 | 1212 |
| LMArena French | 1396 | 1224 |
| LMArena German | 1348 | 1222 |
| LMArena Japanese | 1268 | 1158 |
| LMArena Korean | 1269 | 1151 |
| LMArena Russian | 1303 | 1209 |
| LMArena Spanish | 1376 | 1234 |
Instruction Following Nemotron 3 Nano 30B A3B leads
Nemotron 3 Nano 30B A3B: 68.8 (#180), Phi-4: 60.4 (#251)
| Benchmark | Nemotron 3 Nano 30B A3B | Phi-4 |
|---|---|---|
| LMArena Instruction Following | 1304 | 1201 |
| LiveBench Instruction Following | — | 58.4% |
Long Context Nemotron 3 Nano 30B A3B leads
Nemotron 3 Nano 30B A3B: 39.5 (#175), Phi-4: 36.9 (#226)
| Benchmark | Nemotron 3 Nano 30B A3B | Phi-4 |
|---|---|---|
| LMArena Longer Query | 1300 | 1217 |
Writing & Preference Nemotron 3 Nano 30B A3B leads
Nemotron 3 Nano 30B A3B: 52.3 (#174), Phi-4: 40.5 (#244)
| Benchmark | Nemotron 3 Nano 30B A3B | Phi-4 |
|---|---|---|
| LMArena Text | 1348 | 1217 |
| LMArena Creative Writing | 1269 | 1182 |
| LMArena Multi-Turn | 1319 | 1206 |
| Short-Story Creative Writing | — | 62.6% |
| LiveBench Language | — | 25.6% |
Frequently asked questions
Is Nemotron 3 Nano 30B A3B better than Phi-4?
Nemotron 3 Nano 30B A3B is the stronger model overall, scoring 40.6 to 31.2 on the Noometry Index.
Which is cheaper, Nemotron 3 Nano 30B A3B or Phi-4?
Phi-4 is cheaper. It lists at $0.07 per million input tokens and $0.14 per million output tokens; Nemotron 3 Nano 30B A3B lists at $0.05 and $0.20.
Is Nemotron 3 Nano 30B A3B or Phi-4 better for coding?
Nemotron 3 Nano 30B A3B scores higher on coding benchmarks: 40.5 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
Nemotron 3 Nano 30B A3B does, with 262K tokens against 128K.
How many benchmarks do Nemotron 3 Nano 30B A3B and Phi-4 share?
18 benchmarks have published results for both models. Nemotron 3 Nano 30B A3B has 18 scored results on Noometry and Phi-4 has 37.