DeepSeek-V4.1-Flash-1017-oQ4e-mtp

DeepSeek-V4.1-Flash-1017-oQ4e-mtp

A mixed-precision MLX copy of DeepSeek-V4.1-Flash-1017, a huge multimodal mixture-of-experts model. Made for people with hundreds of GB of memory who want to run it with MLX tools.

Made by mlx-community on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face.

About

What it is: This is a mixed-precision MLX version of DeepSeek-V4.1-Flash-1017 from deepseek-ai. The original is a small update that improves reasoning, coding, agent use and image understanding. It is a multimodal mixture of experts with 552B parameters in its backbone. The oQ method from oMLX version 0.7.1 quantized it, and the whole checkpoint averages 4.529 bits per weight.

Who it suits: It suits advanced users and teams with huge amounts of memory who want to run a big multimodal model through MLX instead of a hosted service.

Industry | Who uses it | How they use it | Example
AI infrastructure | Platform engineers | Host a huge open model on MLX hardware | Serving the model to internal tools
Research | ML researchers | Study mixed-precision quantization of a huge model | Comparing the bit distribution with other copies
Software | Advanced developers | Test coding and agentic behavior locally | Trying a coding task with the model
Multimodal work | Applied teams | Combine text and image inputs in one model | Asking a question about a picture

Skip it if you have a normal laptop or desktop, since the stored tensors take about 402 GiB.

Where to use it:

• MLX based tools
• Machines with hundreds of GB of memory
• Windows, macOS and Linux, with Python, to download the files

Use case | Who | What they get back
Reasoning and coding | Developers | Text replies from the full model
Image and text input | Applied teams | Answers that use pictures
Quantization study | Researchers | A reference mixed-precision checkpoint

How to set it up:

• Python, to run the download command
• About 402 GiB of storage for the tensors, according to the card
• MLX tooling that can load the checkpoint

1. Install the Hugging Face tools.
2. Download the model files to a drive with plenty of space.
3. Load the checkpoint with your MLX tooling.

pip install -U huggingface_hub
hf download mlx-community/DeepSeek-V4.1-Flash-1017-oQ4e-mtp

How to use it: The part of the card available here gives no run commands, only how the model was quantized. Load the checkpoint in MLX tooling that supports it, and see the original DeepSeek model for how to prompt it.

Component | Storage
Language backbone | 300.132 GB
Engram tables | 122.884 GB
Vision encoder | 0.971 GB
MTP head | 8.047 GB

Example: You load the checkpoint in an MLX tool on a machine with enough memory and ask a coding question. The model answers with text. The card does not show a sample run, so treat this as the general idea.

For newcomers and experts: This is a giant model squeezed into a smaller file, and it is meant for people with huge amounts of memory. If you are new, start with a small model first and come back to this one later.

If you know the field, oQ4e leaves the projections at their FP4 and FP8 precision and shrinks only the Engram tables, using 4-bit affine storage in groups of 32. The base model uses compressed sparse attention, Engram conditional memory and DSpark speculative decoding. Without the Engram tables the average is 4.365 bits per weight.

Start from DeepSeek-V4.1-Flash-1017
Keep FP4 and FP8 projections
Quantize the Engram tables to 4-bit
Save the MLX checkpoint

Title: Storage by component, in GB
Language backbone: 300.132 GB
Engram tables: 122.884 GB
Vision encoder: 0.971 GB
MTP head: 8.047 GB

Good to know:

• Size: 763.205B logical weights in total, with about 402.363 GiB of tensor storage.
• Contents: it bundles the text model, the Engram lookup tables, a vision encoder, an aligner that joins vision to text, and the speculative decoding weights.
• Source: a quantized copy of DeepSeek-V4.1-Flash-1017.
• License: MIT, following the source model.
• Cost to run: none named, though the memory needs are huge.
• Data it reaches: it downloads model files from Hugging Face and works on the text and images you give it.
• Not for: ordinary laptops or small desktops.

On Hugging Face, 13 November 2026

• Task: image text to text
• Library: mlx
• Parameters: 129B
• Weights: about 516 GB in U32
• Formats: Safetensors
• Downloads in the last 30 days: 8,491
• Likes: 21
• License: mit
• Based on: deepseek-ai/DeepSeek-V4.1-Flash-1017
• Tags: safetensors, deepseek_v41, quantized, image-text-to-text

Links

• Model: https://huggingface.co/mlx-community/DeepSeek-V4.1-Flash-1017-oQ4e-mtp
• Files: https://huggingface.co/mlx-community/DeepSeek-V4.1-Flash-1017-oQ4e-mtp/tree/main

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download mlx-community/DeepSeek-V4.1-Flash-1017-oQ4e-mtp

Then follow the usage in its model card. Nothing in this kit runs by itself.