Model comparison
GPT-5.4 nano vs Nemotron 3 Super
GPT-5.4 nano is the stronger model overall, scoring 41.9 to 40.1 on the Noometry Index. Nemotron 3 Super costs 2.7× less per token, which makes it the better buy when GPT-5.4 nano's lead doesn't matter for your workload.
Last verified . 19 shared benchmarks.
NVIDIASummary
- They share 19 benchmarks with published results for both. GPT-5.4 nano scores higher in 7 categories and Nemotron 3 Super in 1 category; 4 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.4 nano leads 23.7 to 18.7.
- The biggest single-benchmark swing is WeirdML: 49.2% for GPT-5.4 nano and 38% for Nemotron 3 Super.
- Nemotron 3 Super is cheaper at $0.08 / $0.45 per million input/output tokens, against $0.20 / $1.25 for GPT-5.4 nano.
- GPT-5.4 nano accepts more context: 400K tokens versus 262K.
- Nemotron 3 Super has downloadable open weights; the other is API-only.
Side by side
| GPT-5.4 nano | Nemotron 3 Super | |
|---|---|---|
| Provider | OpenAI | NVIDIA |
| Noometry Index | 41.9 | 40.1 |
| Released | 2026-03-17 | 2026-03-11 |
| Weights | Proprietary | Open |
| Context window | 400K | 262K |
| Max output | 128K | 131K |
| Input $ / M tokens | $0.20 | $0.08 |
| Output $ / M tokens | $1.25 | $0.45 |
| Results tracked | 40 | 21 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.4 nano leads
GPT-5.4 nano: 43.6 (#84), Nemotron 3 Super: 39.4 (#158)
| Benchmark | GPT-5.4 nano | Nemotron 3 Super |
|---|---|---|
| SciCode | 46.9% | 36% |
| WeirdML | 49.2% | 38% |
| LMArena Coding | 1405 | 1403 |
| ALE-Bench | 1,005 | 213.9 |
Reasoning GPT-5.4 nano leads
GPT-5.4 nano: 23.7 (#173), Nemotron 3 Super: 18.7 (#276)
| Benchmark | GPT-5.4 nano | Nemotron 3 Super |
|---|---|---|
| CritPt | 9.3% | 3.1% |
| LMArena Hard Prompts | 1381 | 1386 |
| ARC-AGI-2 | 5.7% | — |
| Kagi LLM Benchmark | 39.7% | — |
| NYT Connections (extended) | — | 15.4% |
| ARC-AGI-1 | 51.5% | — |
| Chess Puzzles | 30% | — |
| Mystery Game Puzzles | 9% | — |
| DTBench | 80.3% | — |
| LMCA | 36.9% | — |
| Epoch Capabilities Index | 145.81 | — |
| ForecastBench | 57.3 | — |
Math GPT-5.4 nano leads
GPT-5.4 nano: 40.9 (#88), Nemotron 3 Super: 39.6 (#101)
| Benchmark | GPT-5.4 nano | Nemotron 3 Super |
|---|---|---|
| LMArena Math | 1406 | 1380 |
| FrontierMath (Tiers 1-3) | 44.9% | — |
| FrontierMath Tier 4 | 12.2% | — |
| MathArena Final-Answer Competitions | — | 60.4% |
| OTIS Mock AIME 2024-2025 | 87.8% | — |
| ProofBench | 5% | — |
| FrontierMath (Feb 2025 set) | 25.9% | — |
| FrontierMath Tier 4 (v1) | 6.3% | — |
Knowledge GPT-5.4 nano leads
GPT-5.4 nano: 41.9 (#103), Nemotron 3 Super: 38.9 (#139)
| Benchmark | GPT-5.4 nano | Nemotron 3 Super |
|---|---|---|
| LMArena Expert | 1396 | 1398 |
| GPQA Diamond | 78.5% | — |
| SimpleQA Verified | 11.7% | — |
| Vectara Hallucination Rate | 3.1% | — |
Multimodal Not comparable
GPT-5.4 nano: 36.7 (#78), Nemotron 3 Super: —
| Benchmark | GPT-5.4 nano | Nemotron 3 Super |
|---|---|---|
| LMArena Vision | 1196 | — |
Multilingual Too close to call
GPT-5.4 nano: 48.6 (#140), Nemotron 3 Super: 48.4 (#144)
| Benchmark | GPT-5.4 nano | Nemotron 3 Super |
|---|---|---|
| LMArena Non-English | 1359 | 1355 |
| LMArena Chinese | 1392 | 1432 |
| LMArena French | 1396 | 1405 |
| LMArena German | 1367 | 1341 |
| LMArena Russian | 1363 | 1336 |
| LMArena Spanish | 1371 | 1417 |
| LMArena Japanese | 1343 | — |
| LMArena Korean | 1320 | — |
Instruction Following Too close to call
GPT-5.4 nano: 71.9 (#144), Nemotron 3 Super: 71.2 (#154)
| Benchmark | GPT-5.4 nano | Nemotron 3 Super |
|---|---|---|
| LMArena Instruction Following | 1362 | 1347 |
Long Context Too close to call
GPT-5.4 nano: 41.6 (#137), Nemotron 3 Super: 41.5 (#139)
| Benchmark | GPT-5.4 nano | Nemotron 3 Super |
|---|---|---|
| LMArena Longer Query | 1366 | 1362 |
Writing & Preference Too close to call
GPT-5.4 nano: 55.7 (#142), Nemotron 3 Super: 55.9 (#140)
| Benchmark | GPT-5.4 nano | Nemotron 3 Super |
|---|---|---|
| LMArena Text | 1372 | 1378 |
| LMArena Creative Writing | 1314 | 1317 |
| LMArena Multi-Turn | 1382 | 1369 |
Frequently asked questions
Is GPT-5.4 nano better than Nemotron 3 Super?
GPT-5.4 nano is the stronger model overall, scoring 41.9 to 40.1 on the Noometry Index. Nemotron 3 Super costs 2.7× less per token, which makes it the better buy when GPT-5.4 nano's lead doesn't matter for your workload.
Which is cheaper, GPT-5.4 nano or Nemotron 3 Super?
Nemotron 3 Super is cheaper. It lists at $0.08 per million input tokens and $0.45 per million output tokens; GPT-5.4 nano lists at $0.20 and $1.25.
Is GPT-5.4 nano or Nemotron 3 Super better for coding?
GPT-5.4 nano scores higher on coding benchmarks: 43.6 versus 39.4 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 nano does, with 400K tokens against 262K.
How many benchmarks do GPT-5.4 nano and Nemotron 3 Super share?
19 benchmarks have published results for both models. GPT-5.4 nano has 40 scored results on Noometry and Nemotron 3 Super has 21.