Huihui Qwen3.6 35B A3B Abliterated needs about 21.0 GB of memory at MXFP4_MOE (a 19.8 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 2 quants, from 19.8 GB (MXFP4_MOE) to 36.9 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 2.7 GB.
Huihui Qwen3.6 35B A3B Abliterated – large multimodal LLM with reduced safety filtering
This model is a 35B parameter Qwen3.6 variant with 3B active experts. It supports vision and a native context length of 262,144 tokens, extendable to 1,010,000 tokens. The model is released under Apache‑2.0 and is intended for research and controlled use.
Good at
- Large context handling
- Vision‑enabled multimodal tasks
- Mixture‑of‑Experts efficiency
Limits
- High VRAM requirement – 19.79 GB file (MXFP4_MOE) or 36.9 GB (Q8_0)
- Reduced safety filtering – risk of sensitive content
- Not suitable for production or public deployment
Best for
- Research on multimodal LLMs
- Testing vision‑enabled inference
- Exploring Mixture‑of‑Experts behavior
Who it’s for. Developers and researchers who need a large, vision‑capable LLM for experimentation. The model’s safety filtering is minimal, so it should only be used in controlled environments with active monitoring.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| MXFP4_MOE | 19.8 GB | 21.0 GB | Too big | Get file → | |
| Q8_0 | 36.9 GB | 38.1 GB | Too big | Get file → |
For a single RTX 3090 or 4090, use the MXFP4_MOE quant – file size 19.79 GB and estimated VRAM usage 21 GB at 8 k context. If you have more VRAM, the Q8_0 quant (36.9 GB, 38.1 GB at 8 k context) offers full quality. Memory estimates are approximate.
What GPU do you need for Huihui Qwen3.6 35B A3B Abliterated?
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 MXFP4_MOE (19.8 GB file, about 21.0 GB in use).
- 32 GB VRAM: the best quant that fits is MXFP4_MOE (19.8 GB file, about 21.0 GB in use).
- 48 GB VRAM: the best quant that fits is Q8_0 (36.9 GB file, about 38.1 GB in use).
Which Huihui Qwen3.6 35B A3B Abliterated quant should you choose?
MXFP4_MOE (19.8 GB) is the middle option in this repository. Lower bit counts mean smaller files and more quality loss.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Huihui Qwen3.6 35B A3B Abliterated
How much VRAM does Huihui Qwen3.6 35B A3B Abliterated need?
About 21.0 GB at MXFP4_MOE with an 8K context window. That is the 19.8 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 Huihui Qwen3.6 35B A3B Abliterated on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 21.0 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Huihui Qwen3.6 35B A3B Abliterated GGUF?
It depends on the quant: MXFP4_MOE is 19.8 GB, Q8_0 is 36.9 GB.
What context length does Huihui Qwen3.6 35B A3B Abliterated support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.7 GB at 8K and 2.7 GB at 32K.
Where can I download Huihui Qwen3.6 35B A3B Abliterated GGUF?
The files are on Hugging Face in the andyjack/Huihui-Qwen3.6-35B-A3B-abliterated-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 Huihui Qwen3.6 35B A3B Abliterated 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?
No, the smallest quant requires at least 19.79 GB of VRAM.
Does it support vision?
Yes, the model includes a vision encoder.
Is it safe for public use?
No, safety filtering is minimal; use only in research or controlled settings.