LLM
GA
Open weights
Nemotron 3 Super
Enterprise self-hosted agentic workloads
Release2026-03
Context1M
Max output—
Parameters120B
Local / VRAM140–192 GB
LicenseOpen weights
Overview
Nemotron 3 Super is a model in the LLM category from NVIDIA. It was released in 2026-03. The vendor is based in US. Best for: Enterprise self-hosted agentic workloads.
Hardware: FP8 ~120 GB; 2x H100 80 GB or 1x B200.
Strengths
- 1M context
- Open weights
- Strong tool use
Limits
- Large hardware
- Limited community ecosystem
Specifications
| Vendor | NVIDIA |
|---|---|
| Category | LLM |
| Release | 2026-03 |
| Status | GA |
| License | Open weights |
| License name | NVIDIA Open Model License |
| Commercial use | Allowed |
| Parameters | 120B |
| Active parameters | 12B |
| Architecture | MoE transformer |
| Context window | 1M |
| Input modalities | Text, Image |
| Output modalities | Text |
| Prompt types | Text |
| Negative prompts | No |
| Max outputs | 1 |
| Local deployment | Yes |
| VRAM | 140–192 GB |
| System RAM | 192–384 GB |
| API access | Yes |
| Pricing | Free weights |
| Confidence | Medium |
| Vendor country | US |
| Vendor type | Open lab |
Pricing & hardware
Free weights
This model can be self-hosted or run locally. Approximate GPU memory: 140–192 GB. Approximate system RAM: 192–384 GB. API access is available.