DeepSeek DQ3 K M needs about 342.2 GB of memory at its main file (a 300.8 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 2 quants, from 300.8 GB to 300.8 GB. Longer context needs more: at 32K tokens the KV cache alone adds about 163.7 GB.
DeepSeek DQ3 K M is a 671B-parameter DeepSeek model in GGUF format, quantized by UnicomAI. Its trained context length is 160K tokens. The repository has 2 quant files, from 300.8 GB to 300.8 GB.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 300.8 GB (7 parts) | 342.2 GB | Too big | Get file → | |
| — | 300.8 GB (7 parts) | 342.2 GB | Too big | Get file → |
What GPU do you need for DeepSeek DQ3 K M?
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.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about DeepSeek DQ3 K M
How much VRAM does DeepSeek DQ3 K M need?
About 342.2 GB at its main file with an 8K context window. That is the 300.8 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 DQ3 K M on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 342.2 GB, so you would need a larger GPU or CPU offloading.
What is the file size of DeepSeek DQ3 K M GGUF?
It depends on the quant: file is 300.8 GB, file is 300.8 GB.
What context length does DeepSeek DQ3 K M support?
The GGUF header lists a trained context length of 163,840 tokens (160K). The KV cache grows with the context you use: about 40.9 GB at 8K and 163.7 GB at 32K.
Where can I download DeepSeek DQ3 K M GGUF?
The files are on Hugging Face in the UnicomAI/DeepSeek-DQ3_K_M 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 DQ3 K M use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.