Qwen3.8 27B TQ needs about 13.2 GB of memory at its main file (a 12.5 GB file) with an 8K context window, so it fits on a 16 GB GPU. This repository has 5 quants, from 0.8 GB to 14.7 GB. Longer context needs more: at 32K tokens the KV cache alone adds about 0.7 GB.
Qwen3.8-27B TQ – a 27B dense vision‑language model with speculative decoding
Qwen3.8-27B TQ is a 27.3B parameter dense model in the Qwen family. It supports vision‑language tasks and long context up to 262144 tokens. The model is available in four GGUF sizes from 11.85 GB to 14.72 GB.
Good at
- Vision‑language understanding
- Long‑horizon agentic tasks
- Coding and professional work
Limits
- No native support for stock llama.cpp, Ollama or LM Studio
- Requires tqLLM engine for TQ tensors
- Large memory requirement for 8 k context
Best for
- Deploying on 16 GB GPUs with TQ_S or TQ_XS
- Running exact speculative decoding with DFlash2
- Serving OpenAI‑compatible chat endpoints
Who it’s for. Developers needing a dense vision‑language model that can run on consumer GPUs with speculative decoding. The model is suitable for teams that can use the tqLLM engine and have access to 16–24 GB GPU memory.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.8 GB | 1.5 GB | Fits | Get file → | |
| — | 11.9 GB | 12.5 GB | Too big | Get file → | |
| — | 12.5 GB | 13.2 GB | Too big | Get file → | |
| — | 13.1 GB | 13.8 GB | Too big | Get file → | |
| — | 14.7 GB | 15.4 GB | Too big | Get file → |
For a 16 GB GPU, the 11.85 GB TQ_S file is recommended; it uses about 12.5 GB at 8 k context. If 24 GB GPUs are available, the 14.72 GB TQ_L file can be used, estimating 15.4 GB at 8 k context. The 13.14 GB TQ_M file is a middle ground for 16 GB GPUs, using about 13.8 GB at 8 k context.
What GPU do you need for Qwen3.8 27B TQ?
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 the main file (0.8 GB file, about 1.5 GB in use).
- 6 GB VRAM: the best quant that fits is the main file (0.8 GB file, about 1.5 GB in use).
- 8 GB VRAM: the best quant that fits is the main file (0.8 GB file, about 1.5 GB in use).
- 12 GB VRAM: the best quant that fits is the main file (0.8 GB file, about 1.5 GB in use).
- 16 GB VRAM: the best quant that fits is the main file (14.7 GB file, about 15.4 GB in use).
Which Qwen3.8 27B TQ quant should you choose?
If memory is tight, the smallest file is the first one in the table at 0.8 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
Source
Frequently asked questions about Qwen3.8 27B TQ
How much VRAM does Qwen3.8 27B TQ need?
About 13.2 GB at its main file with an 8K context window. That is the 12.5 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 TQ on an 8 GB GPU?
Yes, with the smaller files: the 0.8 GB file needs about 1.5 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Qwen3.8 27B TQ GGUF?
It depends on the quant: file is 0.8 GB, file is 11.9 GB, file is 12.5 GB, file is 13.1 GB, file is 14.7 GB.
What context length does Qwen3.8 27B TQ 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 TQ GGUF?
The files are on Hugging Face in the YTan2000/Qwen3.8-27B-TQ 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 TQ 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 TQ on a 16 GB GPU?
Yes, the TQ_S (11.85 GB) and TQ_M (13.14 GB) files fit on 16 GB GPUs, using approximately 12.5 GB and 13.8 GB at 8 k context respectively.
Does this model support vision tasks?
Yes, it is a native vision‑language model that understands images and videos.
What engine is required to run the TQ tensors?
The model requires the tqLLM engine; stock llama.cpp, Ollama, or LM Studio cannot load the TQ tensors.