DE

DeepSeek V3.2 light

DeepSeekGGUF by sszymczykUpdated 2026-03-22
Download on Hugging Face
Parameters672B
Context160K
Layers62
Architecturedeepseek32
Smallest quant404.1 GB
Quick answer

DeepSeek V3.2 light needs about 405.7 GB of memory at Q4_K_M (a 404.1 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 2 quants, from 404.1 GB (Q4_K_M) to 714.3 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 4.4 GB.

DeepSeek V3.2 light is a 672B-parameter DeepSeek model in GGUF format, quantized by sszymczyk. Its trained context length is 160K tokens. The repository has 2 quant files, from 404.1 GB (Q4_K_M) to 714.3 GB (Q8_0).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K_M404.1 GB (31 parts) 405.7 GB
Too big Get file →
Q8_0714.3 GB (31 parts) 715.9 GB
Too big Get file →

What GPU do you need for DeepSeek V3.2 light?

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 V3.2 light quant should you choose?

Q4_K_M is the usual starting point for DeepSeek V3.2 light: it balances file size and quality well. Here it is a 404.1 GB file.

The Q8_0 file is 714.3 GB, about 1.8 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturedeepseek32
Parameters672B (approximate, mixture of experts)
Layers62
Trained context163,840 tokens
KV cache (fp16)132 KB per token

Source

Quantized bysszymczyk
FamilyDeepSeek
Licensemit
Downloads (HF)17
Last updated2026-03-22

Frequently asked questions about DeepSeek V3.2 light

How much VRAM does DeepSeek V3.2 light need?

About 405.7 GB at Q4_K_M with an 8K context window. That is the 404.1 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 V3.2 light on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 405.7 GB, so you would need a larger GPU or CPU offloading.

What is the file size of DeepSeek V3.2 light GGUF?

It depends on the quant: Q4_K_M is 404.1 GB, Q8_0 is 714.3 GB.

What context length does DeepSeek V3.2 light support?

The GGUF header lists a trained context length of 163,840 tokens (160K). The KV cache grows with the context you use: about 1.1 GB at 8K and 4.4 GB at 32K.

Where can I download DeepSeek V3.2 light GGUF?

The files are on Hugging Face in the sszymczyk/DeepSeek-V3.2-light-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 V3.2 light 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.

Similar models

Browse all →

More from sszymczyk

All by sszymczyk →

More DeepSeek models

All DeepSeek →

Other large models

All large models →