All tasks / Models
Nemotron 3.5 Lightning 30B-A3B
NVIDIA's small open-weights Nemotron MoE (OpenMDW-1.1), 30B total and 3B active, runs on one consumer GPU.
- NVIDIA
- Open weights
- Released 1 Aug 2026
Where it ranks
- BChat & assistantsModels · Local · 1,351 on LMArena Text Arena (overall, style control)
- CCodingModels · Local · 13 on Artificial Analysis Intelligence Index (secondary benchmark)
Facts
- Vendor
- NVIDIA
- Family
- Nemotron Lightning
- Released
- 1 Aug 2026
- Weights
- Open: download and run it yourself
- License
- OpenMDW-1.1
- Parameters
- 31.6B (3B active)
- GPU memory to run locally
- ~20 GB (4-bit ≈ 31.6 B × 0.57 GB + 1.5 GB ≈ 20 GB, estimate.)
- API price
- $0.07 in / $0.2 out per 1M tokens (OpenRouter listed price)
- Access
- Free (open weights)
- Last verified
- 19 Oct 2026
Price from openrouter.ai, 19 Oct 2026.
Benchmark results
| Benchmark | Result | Entry on the leaderboard | As of | Source |
|---|---|---|---|---|
| Artificial Analysis Intelligence Index | 13 | Nemotron 3.5 LightningRaw value from the page's embedded model data; the site displays it rounded to an integer. | 19 Oct 2026 | artificialanalysis.ai |
| LMArena Text Arena (overall, style control) | 1,351 | nvidia-nemotron-3.5-lightning-30b-a3b-nvfp4From LMArena's official leaderboard-dataset (text_style_control, split latest, category overall, published 2026-10-15), which mirrors https://arena.ai/leaderboard/text. Rank 216; 19938 votes; 95% CI 1345.8-1356.9. | 15 Oct 2026 | huggingface.co |
Where you can use it
No tracked tool lists this model yet; as open weights it runs in local runtimes.