---
title: DeepSeek-V4.1-Flash-1017-MLXe
canonical_url: "https://www.modelscope.ai/models/inferencerlabs/DeepSeek-V4.1-Flash-1017-MLXe"
md_url: "https://www.modelscope.ai/models/inferencerlabs/DeepSeek-V4.1-Flash-1017-MLXe.md"
repository: inferencerlabs/DeepSeek-V4.1-Flash-1017-MLXe
last_updated: 2026-10-23
pipeline_tag: image-text-to-text
tasks:
  - image-text-to-text
model_type:
  - deepseek_v41
architectures:
  - DeepseekV41ForCausalLM
base_model:
  - deepseek-ai/DeepSeek-V4.1-Flash-1017
base_model_relation: quantized
parameters: 143.9B
tensor_type:
  - BF16
  - F32
  - U32
  - U8
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
language:
  - en
downloads: 27
stars: 0
tags:
  - quantized
  - mlx
---

# DeepSeek-V4.1-Flash-1017-MLXe

> DeepSeek-V4.1-Flash-1017-MLXe - An open-source model by inferencerlabs on ModelScope. DeepSeek-V4.1-Flash-1017 See DeepSeek-V4.1-Flash-1017 in action: demonstration videos Tested with Inferencer M3 Ultra 512GB: 15.9 tokens/s @ 1000 tokens 278 GiB

inferencerlabs/DeepSeek-V4.1-Flash-1017-MLXe is a 143.9B-parameter image-text-to-text model on ModelScope. derived from deepseek-ai/DeepSeek-V4.1-Flash-1017.

- **Repository**: inferencerlabs/DeepSeek-V4.1-Flash-1017-MLXe
- **Tasks**: image-text-to-text
- **Parameters**: 143.9B
- **Base model**: deepseek-ai/DeepSeek-V4.1-Flash-1017
- **Tags**: quantized, mlx
- **Downloads**: 27
- **Stars**: 0
- **Last updated**: 2026-10-23

Source: https://www.modelscope.ai/models/inferencerlabs/DeepSeek-V4.1-Flash-1017-MLXe

---

# DeepSeek-V4.1-Flash-1017
See DeepSeek-V4.1-Flash-1017 in action: [demonstration videos](https://youtube.com/xcreate)
   
#### Tested with [Inferencer](https://inferencer.com)
- M3 Ultra 512GB: ~15.9 tokens/s @ 1000 tokens ~278 GiB

<p style="margin-bottom:0px;">
This build repacks the language model's existing quantized weights, with the vision layers quantized to 9 bits and the engrams streamed from storage to reduce the memory footprint. This configuration empirically yielded the highest quality in our testing.
</p>

![Screenshot 1](https://cdn-uploads.huggingface.co/production/uploads/688479d616f1ec82fa645019/o6qS3AH3mIv8O4RbWuvTL.jpeg)
![Screenshot 2](https://cdn-uploads.huggingface.co/production/uploads/688479d616f1ec82fa645019/bSLGJNG93dm-eXuzR3kMm.jpeg)

##### 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 [DeepSeek-V4.1-Flash-1017](https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017).

## 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.