QW

ThinkingCap Qwen3.8 27B

QwenGGUF by bottlecapaiUpdated 2026-10-07
Download on Hugging Face
Parameters27.3B
Context256K
Layers65
Architectureqwen35
Smallest quant14.3 GB
Quick answer

ThinkingCap Qwen3.8 27B needs about 17.0 GB of memory at its main file (a 14.3 GB file) with an 8K context window, so it fits on a 24 GB GPU. Longer context needs more: at 32K tokens the KV cache alone adds about 8.7 GB.

ThinkingCap Qwen3.8 27B – 27‑billion‑parameter LLM with vision and reasoning

The ThinkingCap Qwen3.8 27B model is a 27‑billion‑parameter Qwen architecture fine‑tuned for reasoning and vision tasks. The GGUF file is 14.3 GB and requires an 18 GB GPU to run with image input, MTP speculative decoding, and a 32k‑token context. It matches bf16 performance on all five evaluated benchmarks.

Good at

  • Reasoning
  • Vision input
  • Large context (32k tokens)

Limits

  • Requires 18 GB GPU memory
  • No text‑only mode without vision projector
  • License type is other

Best for

  • Complex reasoning queries
  • Image‑based question answering
  • Long‑form content generation

Who it’s for. Developers and researchers needing a large‑parameter model that supports both text and image inputs. Suitable for applications that require extended context and multi‑token prediction.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—14.3 GB 17.0 GB
Too big Get file →

The provided GGUF is 14.3 GB and needs 18 GB VRAM. If a larger GPU is available, a higher‑precision quant (e.g., IQ4_XS 15.5 GB) can be used for slightly better performance. For 18 GB VRAM, the 14.3 GB file is the recommended choice.

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

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 the main file (14.3 GB file, about 17.0 GB in use).

Specifications

Architecture from the GGUF header

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

Source

Quantized bybottlecapai
FamilyQwen
Licenseother
Downloads (HF)10,222
Last updated2026-10-07

Frequently asked questions about ThinkingCap Qwen3.8 27B

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

About 17.0 GB at its main file with an 8K context window. That is the 14.3 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 ThinkingCap Qwen3.8 27B on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 17.0 GB, so you would need a larger GPU or CPU offloading.

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

It depends on the quant: file is 14.3 GB.

What context length does ThinkingCap Qwen3.8 27B 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 ThinkingCap Qwen3.8 27B GGUF?

The files are on Hugging Face in the bottlecapai/ThinkingCap-Qwen3.8-27B-GGUF-RCO-4.2bpw repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does ThinkingCap Qwen3.8 27B 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 this model on 8 GB VRAM?

No. The 14.3 GB GGUF requires at least 18 GB VRAM to run with vision and a 32k context.

Does it support text‑only inference?

The GGUF includes a vision projector; text‑only mode is not supported without it.

What GPU memory is needed for 32k context?

An 18 GB GPU is required to fit the model, vision projector, and 32k context on the card.

Other GGUFs of ThinkingCap-Qwen3.8-27B

Similar models

Browse all →

More from bottlecapai

All by bottlecapai →

More Qwen models

All Qwen →

Other medium models

All medium models →