DeepSeek V4 Flash 0731 reap 200b needs about 91.7 GB of memory at Q2_K (a 81.4 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 6 quants, from 0.3 GB to 111.5 GB (MXFP4_MOE).
DeepSeek V4 Flash 200B with 178 routed experts, 200B parameters
DeepSeek V4 Flash 200B is a 200‑billion‑parameter model with 178 routed experts per layer after REAP pruning. The GGUF build removes multi‑token‑prediction modules and retains default thinking and sampling settings. It is released under the MIT license.
Good at
- Reasoning with default thinking
- Standard generation with sampling
- Supports 8‑k context length
Limits
- No multi‑token‑prediction modules (no MTP speculative decoding)
- Requires 8‑k context for memory estimates
- Large memory footprint for higher quants
Best for
- Chat applications needing reasoning
- Tool‑calling with DeepSeek prompt format
- LLM inference on llama.cpp
Who it’s for. Engineers and researchers deploying DeepSeek V4 Flash on local hardware. Suitable for users who need a large‑parameter model with routed experts and who can manage the memory requirements of the GGUF files.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.3 GB | 0.9 GB | Fits | Get file → | |
| IQ1_M | 60.5 GB | 68.3 GB | Too big | Get file → | |
| Q2_K | 81.4 GB | 91.7 GB | Too big | Get file → | |
| IQ3_XXS | 87.6 GB | 98.6 GB | Too big | Get file → | |
| Q3_K_M | 92.9 GB | 104.6 GB | Too big | Get file → | |
| MXFP4_MOE | 111.5 GB | 125.4 GB | Too big | Get file → |
For 8‑GB VRAM, use the 0.33 GB unknown quant (~0.9 GB RAM). For 16‑GB VRAM, IQ1_M (60.52 GB, ~68.3 GB RAM) is too large; consider Q2_K (81.41 GB, ~91.7 GB RAM) or IQ3_XXS (87.62 GB, ~98.6 GB RAM). The MXFP4_MOE baseline is 111.51 GB (~125.4 GB RAM).
What GPU do you need for DeepSeek V4 Flash 0731 reap 200b?
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: the best quant that fits is the main file (0.3 GB file, about 0.9 GB in use).
- 6 GB VRAM: the best quant that fits is the main file (0.3 GB file, about 0.9 GB in use).
- 8 GB VRAM: the best quant that fits is the main file (0.3 GB file, about 0.9 GB in use).
- 12 GB VRAM: the best quant that fits is the main file (0.3 GB file, about 0.9 GB in use).
- 16 GB VRAM: the best quant that fits is the main file (0.3 GB file, about 0.9 GB in use).
- 24 GB VRAM: the best quant that fits is the main file (0.3 GB file, about 0.9 GB in use).
- 32 GB VRAM: the best quant that fits is the main file (0.3 GB file, about 0.9 GB in use).
- 48 GB VRAM: the best quant that fits is the main file (0.3 GB file, about 0.9 GB in use).
Which DeepSeek V4 Flash 0731 reap 200b quant should you choose?
Q2_K (81.4 GB) is the middle option in this repository. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.
If memory is tight, the smallest file is the first one in the table at 0.3 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 reap 200b
How much VRAM does DeepSeek V4 Flash 0731 reap 200b need?
About 91.7 GB at Q2_K with an 8K context window. That is the 81.4 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 reap 200b on an 8 GB GPU?
Yes, with the smaller files: the 0.3 GB file needs about 0.9 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of DeepSeek V4 Flash 0731 reap 200b GGUF?
It depends on the quant: file is 0.3 GB, IQ1_M is 60.5 GB, Q2_K is 81.4 GB, IQ3_XXS is 87.6 GB, Q3_K_M is 92.9 GB, MXFP4_MOE is 111.5 GB.
Where can I download DeepSeek V4 Flash 0731 reap 200b GGUF?
The files are on Hugging Face in the puwaer/DeepSeek-V4-Flash-0731-reap-200b-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 reap 200b 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?
The smallest GGUF file is 0.33 GB, requiring about 0.9 GB RAM at 8k context, so it can run on 8 GB VRAM.
Does this model support vision?
No, the model has no vision capabilities.
What is the context length supported?
The memory estimates are given for an 8k context length.