Compare / Nemotron 3.5 Lightning vs Nemotron 3 Nano 30B-A3B
Nemotron 3.5 Lightning vs Nemotron 3 Nano 30B-A3B
Updated November 2026
Verdict
Nemotron 3.5 Lightning and Nemotron 3 Nano 30B-A3B are roughly tied on the Small LM Index (63.6 vs 62.6). On the 3 benchmarks both labs report, Nemotron 3.5 Lightning wins 3 and Nemotron 3 Nano 30B-A3B wins 0. Nemotron 3 Nano 30B-A3B is smaller: it needs about 20.3 GB of memory at 4-bit, against 22.4 GB for Nemotron 3.5 Lightning. Nemotron 3.5 Lightning is cheaper through an API: $0.049 per million tokens against $0.10.
Pick Nemotron 3.5 Lightning if…
- you mostly write code
- API cost matters most
Pick Nemotron 3 Nano 30B-A3B if…
- you have limited memory (about 20.3 GB at 4-bit)
- you want the more widely used model (more quantized builds and community support)
Specifications
| Nemotron 3.5 Lightning | Nemotron 3 Nano 30B-A3B | |
|---|---|---|
| Small LM Index | 63.6 (likely 55–71) | 62.6 (likely 59–66) |
| Lab | NVIDIA · United States | NVIDIA · United States |
| Released | 2026-08-11 | 2025-12-15 |
| Parameters | 31.6B (3.6B active, MoE) | 31.6B (3.5B active, MoE) |
| Memory at 4-bit | ~22.4 GB | ~20.3 GB |
| Context window | 1M tokens | 1M tokens |
| Input | text | text |
| Reasoning | Always thinks | Hybrid (thinking on/off) |
| License | OpenMDW-1.1 | NVIDIA Open Model |
| API price / 1M tokens | $0.049 ($0.025 in, $0.12 out) | $0.10 ($0.060 in, $0.24 out) |
| Hugging Face downloads (30 days) | — | 1.7M |
Benchmarks
Higher is better. Bold marks the winner where both labs report a score. ≈ marks the index model's estimate for a score that wasn't reported, ✓ an independent measurement and * a score measured by another lab.
| Benchmark | Nemotron 3.5 Lightning | Nemotron 3 Nano 30B-A3B |
|---|---|---|
| MMLU-Pro Knowledge | 81.9 | 78.3 |
| GPQA Diamond Knowledge | 75.4 | 73.0 |
| Humanity's Last Exam Reasoning | ≈ 13 | 10.6 |
| AIME Math | ≈ 88 | 89.1 |
| LiveCodeBench Coding | ≈ 73 | 68.3 |
| SWE-bench Verified Coding | 51.6 | 38.8 |
| τ²-Bench Agentic | — | 49.0 |
| IFBench Instruction following | — | 71.5 |
Scores are mostly reported by each lab with its own settings, so small gaps are noise. See the methodology. Open in the interactive comparison to add more models.
Questions
Which is better, Nemotron 3.5 Lightning or Nemotron 3 Nano 30B-A3B?
Nemotron 3.5 Lightning and Nemotron 3 Nano 30B-A3B are roughly tied on the Small LM Index (63.6 vs 62.6). On the 3 benchmarks both labs report, Nemotron 3.5 Lightning wins 3 and Nemotron 3 Nano 30B-A3B wins 0.
Which is cheaper to run?
Nemotron 3.5 Lightning is cheaper through an API: $0.049 per million tokens against $0.10. Nemotron 3 Nano 30B-A3B is smaller: it needs about 20.3 GB of memory at 4-bit, against 22.4 GB for Nemotron 3.5 Lightning.
Which has the longer context window?
Both have a 1M-token context window.
Can I use Nemotron 3.5 Lightning and Nemotron 3 Nano 30B-A3B commercially?
Nemotron 3.5 Lightning is released under OpenMDW-1.1 and Nemotron 3 Nano 30B-A3B under NVIDIA Open Model. Apache 2.0 and MIT allow commercial use; other licenses can add conditions, so read the license before shipping a product.