NVIDIA-Nemotron-3.5-Lightning-30B-A3B-mxfp8
mlx-community/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-mxfp8
Text generation · mlx
MXFP8 Open on Hugging Face
Parameters
32B
Weights (est.)
≈ 33.6 GB
Mac memory (est.)
48 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-3.5-Lightning-30B-A3B-mxfp8 --prompt "Hello"
Details
- Base model
- nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 (quantized) · NVIDIA Nemotron 3.5 Lightning 30B A3B BF16 on the radar →
- Vendor
- NVIDIA
- Task
- Text generation
- Quantisation
- MXFP8
- Licence
- openmdw-1.1
- Downloads
- 919 in the last 30 days · 2.2K all time · 0 likes
- Published
- 2026-08-11 · updated 2026-08-11 (3mo ago)
- Languages
- en, es, fr, de, it, ja
- Tags
- nemotron_h, nvidia, pytorch, nemotron-3.5
Other MLX builds of nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| NVIDIA-Nemotron-3.5-Lightning-30B-A3B-DSpark-bf16 | bf16 | ≈ 1.9 GB | 8 GB+ | 174 |
| NVIDIA-Nemotron-3.5-Lightning-30B-A3B-4bit | 4-bit | ≈ 17.8 GB | 24 GB+ | 2.6K |
| NVIDIA-Nemotron-3.5-Lightning-30B-A3B-OptiQ-4bit | 4-bit | ≈ 17.8 GB | 24 GB+ | 807 |
| NVIDIA-Nemotron-3.5-Lightning-30B-A3B-mxfp4 | MXFP4 | ≈ 17.8 GB | 24 GB+ | 534 |
| NVIDIA-Nemotron-3.5-Lightning-30B-A3B-mxfp8 (this one) | MXFP8 | ≈ 33.6 GB | 48 GB+ | 919 |
| NVIDIA-Nemotron-3.5-Lightning-30B-A3B-8bit | 8-bit | ≈ 33.6 GB | 48 GB+ | 439 |
| NVIDIA-Nemotron-3.5-Lightning-30B-A3B-bf16 | bf16 | ≈ 63.2 GB | 96 GB+ | 939 |