QW

Qwen3.8 27B NVFP4

QwenGGUF by akopytkoUpdated 2026-10-04
Download on Hugging Face
Parameters27.3B
Context256K
Layers65
Architectureqwen35
Smallest quant15.7 GB
Quick answer

Qwen3.8 27B NVFP4 needs about 20.5 GB of memory at Q6_K (a 17.8 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 3 quants, from 15.7 GB to 22.9 GB (Q6_K). Longer context needs more: at 32K tokens the KV cache alone adds about 8.7 GB.

Qwen3.8-27B NVFP4 – 27B vision‑language model with 262144 token context

Qwen3.8-27B NVFP4 is a 27.3B parameter dense model that supports vision and language tasks. It offers a 262144 token context length and is optimized for NVIDIA Blackwell GPUs. The model is released under the apache-2.0 license.

Good at

  • Vision‑language understanding
  • Long‑horizon agentic tasks
  • Coding and professional work

Limits

  • Requires NVIDIA Blackwell GPUs
  • No global per‑tensor scales
  • Large VRAM requirement

Best for

  • Image and video analysis
  • Complex multi‑step task planning
  • Research prototyping

Who it’s for. Developers needing a dense vision‑language model for high‑context tasks. Researchers and engineers working on agentic or coding applications who have access to Blackwell GPUs.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—15.7 GB 18.4 GB
Too big Get file →
Q6_K17.8 GB 20.5 GB
Too big Get file →
Q6_K22.9 GB 25.5 GB
Too big Get file →

For 8 GB VRAM, use the Q6_K quant with file size 17.79 GB; it estimates 20.5 GB memory at 8 k context. If you can allocate 12 GB VRAM, the larger Q6_K quant (22.86 GB, 25.5 GB memory) offers higher fidelity. The unknown quant is 15.72 GB and 18.4 GB memory, suitable for moderate VRAM budgets.

What GPU do you need for Qwen3.8 27B NVFP4?

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: the best quant that fits is Q6_K (17.8 GB file, about 20.5 GB in use).
  • 32 GB VRAM: the best quant that fits is Q6_K (22.9 GB file, about 25.5 GB in use).

Which Qwen3.8 27B NVFP4 quant should you choose?

Q6_K (17.8 GB) is the middle option in this repository. Very close to the original quality for most uses, at a clearly larger size than 4-bit.

If memory is tight, the smallest file is the first one in the table at 15.7 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

Architectureqwen35
Parameters27.3B
Layers65
Trained context262,144 tokens
KV cache (fp16)260 KB per token

Source

Quantized byakopytko
FamilyQwen
Licenseapache-2.0
Downloads (HF)2,992
Last updated2026-10-04

Frequently asked questions about Qwen3.8 27B NVFP4

How much VRAM does Qwen3.8 27B NVFP4 need?

About 20.5 GB at Q6_K with an 8K context window. That is the 17.8 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 Qwen3.8 27B NVFP4 on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 18.4 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Qwen3.8 27B NVFP4 GGUF?

It depends on the quant: file is 15.7 GB, Q6_K is 17.8 GB, Q6_K is 22.9 GB.

What context length does Qwen3.8 27B NVFP4 support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 2.2 GB at 8K and 8.7 GB at 32K.

Where can I download Qwen3.8 27B NVFP4 GGUF?

The files are on Hugging Face in the akopytko/Qwen3.8-27B-NVFP4-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 Qwen3.8 27B NVFP4 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.

Can I run this model on 8 GB VRAM?

Yes, the Q6_K quant at 17.79 GB file size uses about 20.5 GB memory at 8 k context, which fits 8 GB VRAM with efficient batching.

Does this model support vision tasks?

Yes, it is a native vision‑language model that understands images and videos.

What license does it use?

The model is released under the apache-2.0 license.

Other GGUFs of Qwen3.8-27B

Similar models

Browse all →

More from akopytko

All by akopytko →

More Qwen models

All Qwen →

Other medium models

All medium models →