LLM
GA
Open weights
Nemotron-H 2
Efficient inference on long sequences
Release2025-11
Context131K
Max output—
Parameters47B
Local / VRAM48–64 GB
LicenseOpen weights
Overview
Nemotron-H 2 is a model in the LLM category from NVIDIA. It was released in 2025-11. The vendor is based in US. Best for: Efficient inference on long sequences.
Strengths
- Hybrid Mamba
- High throughput
- Single GPU
Limits
- No vision
- 131K context
Specifications
| Vendor | NVIDIA |
|---|---|
| Category | LLM |
| Release | 2025-11 |
| Status | GA |
| License | Open weights |
| License name | NVIDIA Open Model License |
| Commercial use | Allowed |
| Parameters | 47B |
| Active parameters | 47B |
| Architecture | hybrid Mamba-Transformer |
| Context window | 131K |
| Input modalities | Text |
| Output modalities | Text |
| Prompt types | Text |
| Negative prompts | No |
| Max outputs | 1 |
| Local deployment | Yes |
| VRAM | 48–64 GB |
| System RAM | 64–128 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: 48–64 GB. Approximate system RAM: 64–128 GB. API access is available.