---
title: NVIDIA-Nemotron-3-Super-120B-A12B-MLX-Q9
canonical_url: "https://www.modelscope.ai/models/inferencerlabs/NVIDIA-Nemotron-3-Super-120B-A12B-MLX-Q9"
md_url: "https://www.modelscope.ai/models/inferencerlabs/NVIDIA-Nemotron-3-Super-120B-A12B-MLX-Q9.md"
repository: inferencerlabs/NVIDIA-Nemotron-3-Super-120B-A12B-MLX-Q9
last_updated: 2026-06-20
pipeline_tag: text-generation
tasks:
  - text-generation
model_type:
  - nemotron_h
architectures:
  - NemotronHForCausalLM
base_model:
  - nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16
base_model_relation: quantized
parameters: 37.8B
tensor_type:
  - BF16
  - F32
  - U32
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
language:
  - en
downloads: 224
stars: 0
tags:
  - quantized
  - mlx
---

# NVIDIA-Nemotron-3-Super-120B-A12B-MLX-Q9

> NVIDIA-Nemotron-3-Super-120B-A12B-MLX-Q9 - An open-source model by inferencerlabs on ModelScope. See NVIDIA-Nemotron-3-Super-120B-A12B MLX in action - demonstration video

inferencerlabs/NVIDIA-Nemotron-3-Super-120B-A12B-MLX-Q9 is a 37.8B-parameter text-generation model on ModelScope. derived from nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16.

- **Repository**: inferencerlabs/NVIDIA-Nemotron-3-Super-120B-A12B-MLX-Q9
- **Tasks**: text-generation
- **Parameters**: 37.8B
- **Base model**: nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16
- **Tags**: quantized, mlx
- **Downloads**: 224
- **Stars**: 0
- **Last updated**: 2026-06-20

Source: https://www.modelscope.ai/models/inferencerlabs/NVIDIA-Nemotron-3-Super-120B-A12B-MLX-Q9

---

**See NVIDIA-Nemotron-3-Super-120B-A12B MLX in action - [demonstration video](https://youtu.be/MzeRCbnOg9Q)**

#### Tested on a M3 Ultra 512GB RAM using [Inferencer app v1.10.6](https://inferencer.com)
- Single inference ~33.7 tokens/s @ 1000 tokens
- Batched inference ~ total tokens/s across five inferences
- Memory usage: ~130 GiB

<p style="margin-bottom:0px;">
<strong>q9bit quant typically achieves near lossless accuracy in our coding test</strong>
</p>
<table style="border-collapse: collapse; text-align:center; margin-top:10px; margin-bottom:0px;">
<thead>
<tr><th>Quantization</th><th>Perplexity</th><th>Token Accuracy</th><th>Missed Divergence</th></tr>
</thead>
<tbody>
<tr><td><strong>q3.5</strong></td><td>168.0</td><td>43.45%</td><td>72.57%</td></tr>
<tr><td><strong>q4.5</strong></td><td>1.33593</td><td>91.65%</td><td>27.61%</td></tr>
<tr><td><strong>q4.8</strong></td><td>1.28125</td><td>93.75%</td><td>21.15%</td></tr>
<tr><td><strong>q5.5</strong></td><td>1.23437</td><td>95.05%</td><td>17.28%</td></tr>
<tr><td><strong>q6.5</strong></td><td>1.21875</td><td>96.95%</td><td>12.03%</td></tr>
<tr><td><strong>q8.5</strong></td><td>1.21093</td><td>97.55%</td><td>10.50%</td></tr>
<tr><td><strong>q9</strong></td><td>1.21093</td><td>97.55%</td><td>10.50%</td></tr>
<tr><td><strong>Base</strong></td><td>1.20312</td><td>100.0%</td><td>0.000%</td></tr>
</tbody>
</table>

##### Quantized with a modified version of [MLX](https://github.com/ml-explore/mlx)
##### For more details see [demonstration video](https://youtu.be/MzeRCbnOg9Q) or visit [NVIDIA-Nemotron-3-Super-120B-A12B-BF16](https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16).

## Disclaimer

We are not the creator, originator, or owner of any model listed. Each model is created and provided by third parties. Models may not always be accurate or contextually appropriate. You are responsible for verifying the information before making important decisions. We are not liable for any damages, losses, or issues arising from its use, including data loss or inaccuracies in AI-generated content.