QW

Qwen3.8 27B Uncensored GP100

QwenGGUF by fallentreeUpdated 2026-09-16
Download on Hugging Face
Parameters4.47B
Context256K
Layers5
Architecturedflash
Smallest quant2.8 GB
Quick answer

Qwen3.8 27B Uncensored GP100 needs about 16.8 GB of memory at Q4_1 (a 16.1 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 3 quants, from 2.8 GB (Q4_1) to 17.7 GB (Q4_1). Longer context needs more: at 32K tokens the KV cache alone adds about 0.7 GB.

Qwen3.8 27B Uncensored GP100 – 4.47B parameter model with vision support

Qwen3.8 27B Uncensored GP100 is a 4.47B parameter dflash architecture model. It supports vision and is released under the apache-2.0 license. The model is available in GGUF format with multiple Q4_1 quantizations.

Good at

  • Vision tasks
  • Uncensored text generation
  • Large context handling

Limits

  • Requires 128k context tokens
  • No GPU other than sm_60 (Tesla P100) for optimal speed
  • No vision support on non‑sm_60 GPUs

Best for

  • Image‑guided text generation
  • Large‑scale inference on Tesla P100
  • Fine‑tuning with LoRA on 4.47B base

Who it’s for. Developers needing a vision‑capable, uncensored LLM for local inference. Suitable for researchers and engineers working with Tesla P100 GPUs and llama.cpp.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_12.8 GB 3.5 GB
Fits Get file →
Q4_116.1 GB 16.8 GB
Too big Get file →
Q4_117.7 GB 18.3 GB
Too big Get file →

For 8 GB VRAM, use the 2.8 GB Q4_1 file (est. 3.5 GB memory at 8 k context). For 16 GB VRAM, the 16.1 GB Q4_1 file (est. 16.8 GB memory) is suitable. For 32 GB VRAM, the 17.65 GB Q4_1 file (est. 18.3 GB memory) provides the largest context.

What GPU do you need for Qwen3.8 27B Uncensored GP100?

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: the best quant that fits is Q4_1 (2.8 GB file, about 3.5 GB in use).
  • 6 GB VRAM: the best quant that fits is Q4_1 (2.8 GB file, about 3.5 GB in use).
  • 8 GB VRAM: the best quant that fits is Q4_1 (2.8 GB file, about 3.5 GB in use).
  • 12 GB VRAM: the best quant that fits is Q4_1 (2.8 GB file, about 3.5 GB in use).
  • 16 GB VRAM: the best quant that fits is Q4_1 (2.8 GB file, about 3.5 GB in use).
  • 24 GB VRAM: the best quant that fits is Q4_1 (17.7 GB file, about 18.3 GB in use).

Which Qwen3.8 27B Uncensored GP100 quant should you choose?

Q4_1 (16.1 GB) is the middle option in this repository. The usual sweet spot between file size and quality, and the quant most people start with.

If memory is tight, the smallest file is Q4_1 at 2.8 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturedflash
Parameters4.47B
Layers5
Trained context262,144 tokens
KV cache (fp16)20 KB per token

Source

Quantized byfallentree
FamilyQwen
Licenseapache-2.0
Downloads (HF)1,360
Last updated2026-09-16

Frequently asked questions about Qwen3.8 27B Uncensored GP100

How much VRAM does Qwen3.8 27B Uncensored GP100 need?

About 16.8 GB at Q4_1 with an 8K context window. That is the 16.1 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 Qwen3.8 27B Uncensored GP100 on an 8 GB GPU?

Yes, with Q4_1: the 2.8 GB file needs about 3.5 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Qwen3.8 27B Uncensored GP100 GGUF?

It depends on the quant: Q4_1 is 2.8 GB, Q4_1 is 16.1 GB, Q4_1 is 17.7 GB.

What context length does Qwen3.8 27B Uncensored GP100 support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.2 GB at 8K and 0.7 GB at 32K.

Where can I download Qwen3.8 27B Uncensored GP100 GGUF?

The files are on Hugging Face in the fallentree/Qwen3.8-27B-Uncensored-GP100-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 Uncensored GP100 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 Qwen3.8 27B Uncensored GP100 on 8 GB VRAM?

Use the 2.8 GB Q4_1 GGUF file; it estimates 3.5 GB VRAM usage at 8 k context.

Does this model support vision?

Yes, the model has detected vision abilities.

What license applies to this model?

The model is released under the apache-2.0 license.

Other GGUFs of Qwen3.8-27B-Uncensored

Similar models

Browse all →

More from fallentree

All by fallentree →

More Qwen models

All Qwen →

Other small models

All small models →