NVIDIA Nemotron 3.5 Super VL 120B-A12B — MLX 4bit (text)

Text conversion of nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16.

Quantization: 4-bit affine, group size 64; embeddings 8-bit/group-64. Output head, routers and latent projections retain checkpoint precision.
Artifact size: 69.541 GB. Recommend 128 GB+ unified memory (measured peak 72.7 GB at 1K context on M3 Ultra; 96 GB Macs may need a raised GPU wired-memory limit and are not yet verified). Allow headroom beyond the weights for prefill, caches, macOS and applications.
Text only: the vision tower, image projector and unused MTP head are omitted.
Declared context limit: 1,048,576 tokens; this does not promise that the full window fits the recommended memory.

Requirements

Requires Rapid-MLX 0.16.2 or newer. The config retains nemotron_h and opts into Rapid's reviewed precision adapter. Stock loaders do not apply that adapter; only Rapid-MLX is validated for this conversion.

rapid-mlx serve rapid-mlx/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-MLX-4bit

Reasoning and XML tool calling use the original tokenizer/chat template. MTP speculative decoding and image input are not enabled.

License

Original model by NVIDIA under OpenMDW 1.1. This conversion inherits the original model's license terms.

Seed this model

Have the files? Submit a torrent.

Submit a torrent →

Package the files with one command

Have the complete files and permission to share them? Replace the example folder path. The script checks the files, makes a torrent and submits its magnet and checksums. Keep your torrent client seeding. Get your community key from your settings, and read the script before running it.

curl -fsSL https://pirateface.co/package.sh | bash -s -- --repo rapid-mlx/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-MLX-4bit ./model-folder