DeepSeek V4 Flash 0731 MXFP4 Q3 K Q2 K mixed needs about 142.7 GB of memory at MXFP4 (a 126.9 GB file) with an 8K context window, which is more than a 48 GB GPU holds.
DeepSeek V4 Flash 0731 MXFP4 Q3 K Q2 K mixed – 284 billion‑parameter LLM
DeepSeek V4 Flash 0731 is a 284 billion‑parameter model with a speculative decoding module. The quantized MXFP4 version weighs 126.93 GB and requires about 142.7 GB of VRAM at 8 k context. It is released under the MIT license.
Good at
- Large‑scale text generation
- Agentic reasoning with speculative decoding
- High‑throughput inference on CPUs and older GPUs
Limits
- Requires 142.7 GB VRAM at 8 k context
- No vision or multimodal support
- No official chat template
Best for
- Running on high‑end GPUs with large VRAM
- Deploying in research environments needing speculative decoding
- Benchmarking LLM performance
Who it’s for. Developers and researchers who need a large, high‑throughput language model and have access to GPUs with at least 142.7 GB VRAM. The model is suitable for experimentation with speculative decoding and agentic tasks.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| MXFP4 | 126.9 GB | 142.7 GB | Too big | Get file → |
The MXFP4 quantization is the only file available. It needs roughly 142.7 GB VRAM at 8 k context. If you have less VRAM, consider using a different quantization not listed here, as no other options are provided.
What GPU do you need for DeepSeek V4 Flash 0731 MXFP4 Q3 K Q2 K mixed?
These figures add the file size, an fp16 KV cache at an 8K context window and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: no quant fits at an 8K context.
- 16 GB VRAM: no quant fits at an 8K context.
- 24 GB VRAM: no quant fits at an 8K context.
- 32 GB VRAM: no quant fits at an 8K context.
- 48 GB VRAM: no quant fits at an 8K context.
Which DeepSeek V4 Flash 0731 MXFP4 Q3 K Q2 K mixed quant should you choose?
MXFP4 (126.9 GB) is the middle option in this repository. Lower bit counts mean smaller files and more quality loss.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about DeepSeek V4 Flash 0731 MXFP4 Q3 K Q2 K mixed
How much VRAM does DeepSeek V4 Flash 0731 MXFP4 Q3 K Q2 K mixed need?
About 142.7 GB at MXFP4 with an 8K context window. That is the 126.9 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run DeepSeek V4 Flash 0731 MXFP4 Q3 K Q2 K mixed on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 142.7 GB, so you would need a larger GPU or CPU offloading.
What is the file size of DeepSeek V4 Flash 0731 MXFP4 Q3 K Q2 K mixed GGUF?
It depends on the quant: MXFP4 is 126.9 GB.
Where can I download DeepSeek V4 Flash 0731 MXFP4 Q3 K Q2 K mixed GGUF?
The files are on Hugging Face in the BahamutRU/DeepSeek-V4-Flash-0731-MXFP4-Q3_K-Q2_K-mixed-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does DeepSeek V4 Flash 0731 MXFP4 Q3 K Q2 K mixed use?
Hugging Face lists the license as mit. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run this model on 8 GB VRAM?
No. The model requires 142.7 GB VRAM at 8 k context, far exceeding 8 GB.
Does this model support vision tasks?
No. The model has no vision or multimodal capabilities.
What license does this model use?
The model is released under the MIT license.