DeepSeek V4 Flash 0731 ROCmFP3 needs about 100.9 GB of memory at its main file (a 99.7 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 3 quants, from 98.3 GB to 102.3 GB. Longer context needs more: at 32K tokens the KV cache alone adds about 2.9 GB.
DeepSeek V4 Flash 0731 ROCmFP3 – 284 billion‑parameter vision‑enabled LLM
DeepSeek V4 Flash 0731 ROCmFP3 is a 284 billion‑parameter model with vision support and a speculative decoding module. It is released under the MIT license and trained on 1,048,576 context tokens. The model is available as a single 98.29 GB GGUF file.
Good at
- Large‑context reasoning
- Vision‑enabled text generation
- Speculative decoding
Limits
- Requires ~99 GB VRAM for 8 k context
- No sidecar files but large memory footprint
- MIT license limits commercial use without attribution
Best for
- High‑capacity chat agents
- Vision‑augmented content creation
- Complex code generation
Who it’s for. Developers building agents that need large context windows and vision input. Researchers testing speculative decoding on AMD hardware. Teams with 128 GB AMD Strix Halo systems or equivalent GPU memory.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 98.3 GB | 99.5 GB | Too big | Get file → | |
| — | 99.7 GB | 100.9 GB | Too big | Get file → | |
| — | 102.3 GB | 103.5 GB | Too big | Get file → |
Use the provided 98.29 GB GGUF file. Estimated VRAM usage is 99.5 GB at 8 k context, approximate. For lower memory budgets, no smaller quantized versions are available in the file list.
What GPU do you need for DeepSeek V4 Flash 0731 ROCmFP3?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) 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 ROCmFP3 quant should you choose?
If memory is tight, the smallest file is the first one in the table at 98.3 GB, with the quality trade-offs that come with fewer bits.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about DeepSeek V4 Flash 0731 ROCmFP3
How much VRAM does DeepSeek V4 Flash 0731 ROCmFP3 need?
About 100.9 GB at its main file with an 8K context window. That is the 99.7 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 ROCmFP3 on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 99.5 GB, so you would need a larger GPU or CPU offloading.
What is the file size of DeepSeek V4 Flash 0731 ROCmFP3 GGUF?
It depends on the quant: file is 98.3 GB, file is 99.7 GB, file is 102.3 GB.
What context length does DeepSeek V4 Flash 0731 ROCmFP3 support?
The GGUF header lists a trained context length of 1,048,576 tokens (1024K). The KV cache grows with the context you use: about 0.7 GB at 8K and 2.9 GB at 32K.
Where can I download DeepSeek V4 Flash 0731 ROCmFP3 GGUF?
The files are on Hugging Face in the Lucebox/DeepSeek-V4-Flash-0731-ROCmFP3 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 ROCmFP3 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 DeepSeek V4 Flash 0731 on 8 GB VRAM?
No. The model requires about 99 GB VRAM at 8 k context, so 8 GB is insufficient.
Does the model support vision?
Yes. The model has detected vision abilities and includes image routing biases.
What license applies to this model?
The model is released under the MIT license.