Qwen3.8-27B GSQ-RCO needs about 15.0 GB of memory at IQ3_S (a 12.3 GB file) with an 8K context window, so it fits on a 16 GB GPU. This repository has 2 quants, from 12.3 GB (IQ3_S) to 12.4 GB (IQ3_S). Longer context needs more: at 32K tokens the KV cache alone adds about 8.7 GB.
Qwen3.8-27B GSQ-RCO – 27.3B parameter dense vision‑language model
Qwen3.8-27B GSQ-RCO is a 27.3B parameter dense model built on the Qwen3.5 architecture. It supports reasoning, tools, and vision tasks. The model is quantized to IQ3_S for efficient deployment.
Good at
- Reasoning and multi‑step tasks
- Tool usage integration
- Vision‑language understanding
Limits
- No vision support in GGUF quantization
- Large VRAM requirement for full context
- License is Apache‑2.0
Best for
- Complex agentic workflows
- Coding assistance
- Research prototyping
Who it’s for. Developers and researchers needing a compact, dense vision‑language model with reasoning and tool capabilities. The model is suitable for environments where memory constraints are moderate and Apache‑2.0 licensing is acceptable.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| IQ3_S | 12.3 GB | 15.0 GB | Too big | Get file → | |
| IQ3_S | 12.4 GB | 15.1 GB | Too big | Get file → |
For 8 GB VRAM, use the IQ3_S quantized file (~12.33 GB). Estimated memory at 8 k context is about 15 GB, so the model will need additional RAM or swap. For tighter budgets, consider a higher‑bit quantization if available.
What GPU do you need for Qwen3.8-27B GSQ-RCO?
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 IQ3_S (12.4 GB file, about 15.1 GB in use).
Which Qwen3.8-27B GSQ-RCO quant should you choose?
IQ3_S (12.3 GB) is the middle option in this repository. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Qwen3.8-27B GSQ-RCO
How much VRAM does Qwen3.8-27B GSQ-RCO need?
About 15.0 GB at IQ3_S with an 8K context window. That is the 12.3 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 GSQ-RCO on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 15.0 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Qwen3.8-27B GSQ-RCO GGUF?
It depends on the quant: IQ3_S is 12.3 GB, IQ3_S is 12.4 GB.
What context length does Qwen3.8-27B GSQ-RCO 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 GSQ-RCO GGUF?
The files are on Hugging Face in the tooltd/Qwen3.8-27B-GSQ-RCO-BuffedMod-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 GSQ-RCO 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?
The IQ3_S quantized file is 12.33 GB and requires roughly 15 GB of memory at 8 k context, so 8 GB VRAM is insufficient.
Does the model support vision tasks?
Yes, the parent model is a native vision‑language model, but the GGUF quantization may limit vision capabilities.
What license applies?
The model is released under the Apache‑2.0 license.