Qwen3.8 27B MIX needs about 15.3 GB of memory at its main file (a 12.6 GB file) with an 8K context window, so it fits on a 16 GB GPU. This repository has 2 quants, from 12.6 GB to 14.1 GB. Longer context needs more: at 32K tokens the KV cache alone adds about 8.7 GB.
Qwen3.8 27B MIX – a dense multimodal model with vision and large context
Qwen3.8 27B MIX is a 27‑billion‑parameter dense model that supports vision and text. It is based on the Qwen3.5 architecture and offers up to 262,144 training context tokens. The model is available in two custom mixed‑tensor GGUF quantizations designed for 16 GB VRAM GPUs.
Good at
- Vision‑language tasks
- Large context reasoning
- Multimodal inference
Limits
- No explicit license details
- Requires 16 GB VRAM for full context
- No separate vision quantization listed
Best for
- Image captioning with long context
- Video understanding with extended prompts
- Complex multimodal dialogue
Who it’s for. Developers building multimodal applications that need large context windows and native vision support. The model is suitable for research and production use where 16 GB VRAM GPUs are available.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 12.6 GB | 15.3 GB | Too big | Get file → | |
| — | 14.1 GB | 16.8 GB | Too big | Get file → |
For 16 GB VRAM GPUs, use the IQ4-MIX quantization (12.59 GB file, ~15.3 GB memory at 8 k context). If more VRAM headroom is needed, choose IQ3-MIX (14.11 GB file, ~16.8 GB memory at 8 k context). Both are approximate estimates.
What GPU do you need for Qwen3.8 27B MIX?
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: the best quant that fits is the main file (12.6 GB file, about 15.3 GB in use).
- 24 GB VRAM: the best quant that fits is the main file (14.1 GB file, about 16.8 GB in use).
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Qwen3.8 27B MIX
How much VRAM does Qwen3.8 27B MIX need?
About 15.3 GB at its main file with an 8K context window. That is the 12.6 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 MIX on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 15.3 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Qwen3.8 27B MIX GGUF?
It depends on the quant: file is 12.6 GB, file is 14.1 GB.
What context length does Qwen3.8 27B MIX 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 MIX GGUF?
The files are on Hugging Face in the vmarcelo/Qwen3.8-27B-MIX_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 MIX use?
Hugging Face lists the license as other. 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 MIX on a 16 GB GPU?
Yes, the IQ4-MIX quantization is designed for 16 GB VRAM and should fit with full context. The IQ3-MIX quantization offers more headroom for larger contexts.
Does the model support vision tasks?
Yes, it includes a vision encoder and can process images and videos natively.
What is the maximum context length?
The model was trained with 262,144 tokens, but practical limits depend on VRAM and quantization.