NVIDIA-Nemotron-3-Super-120B-A12B-8bit
mlx-community/NVIDIA-Nemotron-3-Super-120B-A12B-8bit
Text generation · mlx
8-bit Open on Hugging Face
Parameters
121B
Weights (est.)
≈ 128 GB
Mac memory (est.)
192 GB+
Size is estimated from the parameter count and quantisation. Memory allows a fifth extra for context plus 2 GB for the system; long contexts need more.
Run it
With mlx-lm on an Apple silicon Mac:
pip install mlx-lmmlx_lm.generate --model mlx-community/NVIDIA-Nemotron-3-Super-120B-A12B-8bit --prompt "Hello"
Details
- Base model
- nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16 (quantized)
- Vendor
- NVIDIA
- Task
- Text generation
- Quantisation
- 8-bit
- Licence
- nvidia-nemotron-open-model-license
- Downloads
- 151 in the last 30 days · 3.0K all time · 0 likes
- Published
- 2026-03-30 · updated 2026-03-31 (7mo ago)
- Languages
- en, fr, es, it, de, ja, zh
- Tags
- nemotron_h, nvidia, pytorch, nemotron-3, latent-moe, custom_code
Other MLX builds of nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| NVIDIA-Nemotron-3-Super-120B-A12B-OptiQ-2bit | 2-bit | ≈ 37.9 GB | 48 GB+ | 371 |
| NVIDIA-Nemotron-3-Super-120B-A12B-4bit | 4-bit | ≈ 68.0 GB | 96 GB+ | 513 |
| NVIDIA-Nemotron-3-Super-120B-A12B-5bit | 5-bit | ≈ 83.1 GB | 96 GB+ | 243 |
| Nemotron-3-Super-120B-A12B-MLX-6bit | 6-bit | ≈ 98.1 GB | 128 GB+ | 379 |
| NVIDIA-Nemotron-3-Super-120B-A12B-6bit | 6-bit | ≈ 98.1 GB | 128 GB+ | 181 |
| NVIDIA-Nemotron-3-Super-120B-A12B-8bit (this one) | 8-bit | ≈ 128 GB | 192 GB+ | 151 |