---
title: Inkling-Small-MLX-Q9
canonical_url: "https://www.modelscope.ai/models/inferencerlabs/Inkling-Small-MLX-Q9"
md_url: "https://www.modelscope.ai/models/inferencerlabs/Inkling-Small-MLX-Q9.md"
repository: inferencerlabs/Inkling-Small-MLX-Q9
last_updated: 2026-07-31
pipeline_tag: image-text-to-text
tasks:
  - image-text-to-text
model_type:
  - inkling_mm_model
architectures:
  - InklingForConditionalGeneration
base_model:
  - thinkingmachines/Inkling-Small
base_model_relation: quantized
parameters: 83.9B
tensor_type:
  - BF16
  - U32
  - F32
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
language:
  - en
downloads: 109
stars: 0
tags:
  - quantized
  - mlx
  - conversational
  - image-text-to-text
  - audio-text-to-text
  - moe
---

# Inkling-Small-MLX-Q9

> Inkling-Small-MLX-Q9 - An open-source model by inferencerlabs on ModelScope. Inkling-Small See Inkling-Small in action: demonstration videos Tested with an M3 Ultra 512 GiB using Inferencer app v2.2.2 Multimodal inference: 25 tokens/s @ 278 GiB

inferencerlabs/Inkling-Small-MLX-Q9 is a 83.9B-parameter image-text-to-text model on ModelScope. derived from thinkingmachines/Inkling-Small.

- **Repository**: inferencerlabs/Inkling-Small-MLX-Q9
- **Tasks**: image-text-to-text
- **Parameters**: 83.9B
- **Base model**: thinkingmachines/Inkling-Small
- **Tags**: quantized, mlx, conversational, image-text-to-text, audio-text-to-text, moe
- **Downloads**: 109
- **Stars**: 0
- **Last updated**: 2026-07-31

Source: https://www.modelscope.ai/models/inferencerlabs/Inkling-Small-MLX-Q9

---

# Inkling-Small
See Inkling-Small in action: [demonstration videos](https://youtube.com/xcreate)
   
#### Tested with an M3 Ultra 512 GiB using [Inferencer app v2.2.2](https://inferencer.com)
- Multimodal inference: ~25 tokens/s @ ~278 GiB

<p style="margin-bottom:0px;">
<strong>Q9 achieves near lossless accuracy in our coding test</strong>

![Screenshot](https://cdn-uploads.huggingface.co/production/uploads/688479d616f1ec82fa645019/2rCJPWGHutM0m2RSdbfh6.jpeg)
</p>
<table style="border-collapse: collapse; text-align:center; margin-top:10px; margin-bottom:0px;">
<thead>
<tr><th>Quantization (bpw)</th><th>Perplexity</th><th>Token Accuracy</th><th>Missed Divergence</th><th>Size</th></tr>
</thead>
<tbody>
<tr><td><strong>Q4.5</strong></td><td>1.11169</td><td>94.93%</td><td>24.01%</td><td>165.99 GB</td></tr>
<tr><td><strong>Q6.5</strong></td><td>1.06942</td><td>98.63%</td><td>10.37%</td><td>239.73 GB</td></tr>
<tr><td><strong>Q9</strong></td><td>1.06713</td><td>99.22%</td><td>10.12%</td><td>331.92 GB</td></tr>
<tr><td><strong>Base</strong></td><td>Untested</td><td>100%</td><td>0.000%</td><td>597.59 GB</td></tr>
</tbody>
</table>

##### Quantized with a modified version of [MLX](https://github.com/ml-explore/mlx)
##### For more details see our [demonstration videos](https://youtube.com/xcreate) or visit [thinkingmachines/Inkling-Small](https://huggingface.co/thinkingmachines/Inkling-Small).

## 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.