Qwen3 235B A22B Abliterated needs about 144.2 GB of memory at Q4_K_M (a 142.2 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 24 quants, from 0.5 GB to 193.0 GB (Q6_K). Longer context needs more: at 32K tokens the KV cache alone adds about 6.3 GB.
Abliterated Qwen3 235B with 22B active experts, 40k context
Qwen3 235B A22B Abliterated is a 235‑billion‑parameter model with 22 billion active experts. It uses a 40 k token context window and is released under Apache‑2.0. The model removes the dominant refusal direction, reducing prompt refusals while keeping full capability.
Good at
- Creative writing with fewer refusals
- Large‑context reasoning up to 40k tokens
- High‑quality prose generation
Limits
- Requires large VRAM for full precision
- No vision support in this variant
- Large context limits memory usage
Best for
- Local fiction writing
- Long‑form content generation
- Research on refusal mitigation
Who it’s for. Developers and researchers who need a large, low‑refusal language model for local deployment. Suitable for teams with high‑end GPUs and a need for extended context handling.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 0.5 GB | 2.6 GB | Fits | Get file → | |
| IQ1_S | 47.9 GB | 50.0 GB | Too big | Get file → | |
| IQ1_M | 53.1 GB | 55.2 GB | Too big | Get file → | |
| IQ2_XXS | 61.8 GB | 63.9 GB | Too big | Get file → | |
| IQ2_XS | 68.8 GB | 70.9 GB | Too big | Get file → | |
| IQ2_S | 70.2 GB | 72.3 GB | Too big | Get file → | |
| IQ2_M | 77.2 GB | 79.2 GB | Too big | Get file → | |
| Q2_K_S | 79.8 GB | 81.8 GB | Too big | Get file → | |
| Q2_K | 85.7 GB | 87.8 GB | Too big | Get file → | |
| IQ3_XXS | 90.4 GB | 92.5 GB | Too big | Get file → | |
| IQ3_XS | 96.0 GB | 98.0 GB | Too big | Get file → | |
| Q3_K_S | 101.4 GB | 103.5 GB | Too big | Get file → | |
| IQ3_S | 101.5 GB | 103.5 GB | Too big | Get file → | |
| IQ3_M | 103.1 GB | 105.1 GB | Too big | Get file → | |
| Q3_K_M | 112.4 GB | 114.5 GB | Too big | Get file → | |
| Q3_K_L | 121.8 GB | 123.9 GB | Too big | Get file → | |
| IQ4_XS | 125.3 GB | 127.4 GB | Too big | Get file → | |
| Q4_0 | 133.1 GB | 135.2 GB | Too big | Get file → | |
| Q4_K_S | 133.7 GB | 135.8 GB | Too big | Get file → | |
| Q4_K_M | 142.2 GB | 144.2 GB | Too big | Get file → | |
| Q4_1 | 147.2 GB | 149.3 GB | Too big | Get file → | |
| Q5_K_S | 161.9 GB | 164.0 GB | Too big | Get file → | |
| Q5_K_M | 166.8 GB | 168.9 GB | Too big | Get file → | |
| Q6_K | 193.0 GB | 195.1 GB | Too big | Get file → |
For 8 GB VRAM, use IQ1_S (47.95 GB) or IQ1_M (53.14 GB) and run with 8 k context; memory estimates are approximate. For 16 GB VRAM, IQ2_XXS (61.8 GB) or IQ2_XS (68.82 GB) are suitable. Larger VRAM allows higher‑quality quants such as Q3_K_M (112.45 GB).
What GPU do you need for Qwen3 235B A22B 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: the best quant that fits is the main file (0.5 GB file, about 2.6 GB in use).
- 6 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 2.6 GB in use).
- 8 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 2.6 GB in use).
- 12 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 2.6 GB in use).
- 16 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 2.6 GB in use).
- 24 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 2.6 GB in use).
- 32 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 2.6 GB in use).
- 48 GB VRAM: the best quant that fits is the main file (0.5 GB file, about 2.6 GB in use).
Which Qwen3 235B A22B Abliterated quant should you choose?
Q4_K_M is the usual starting point for Qwen3 235B A22B Abliterated: it balances file size and quality well. Here it is a 142.2 GB file.
If memory is tight, the smallest file is the first one in the table at 0.5 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 235B A22B Abliterated
How much VRAM does Qwen3 235B A22B Abliterated need?
About 144.2 GB at Q4_K_M with an 8K context window. That is the 142.2 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 235B A22B Abliterated on an 8 GB GPU?
Yes, with the smaller files: the 0.5 GB file needs about 2.6 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Qwen3 235B A22B Abliterated GGUF?
It depends on the quant: file is 0.5 GB, IQ1_S is 47.9 GB, IQ1_M is 53.1 GB, IQ2_XXS is 61.8 GB, IQ2_XS is 68.8 GB, IQ2_S is 70.2 GB and 18 more.
What context length does Qwen3 235B A22B Abliterated support?
The GGUF header lists a trained context length of 40,960 tokens (40K). The KV cache grows with the context you use: about 1.6 GB at 8K and 6.3 GB at 32K.
Where can I download Qwen3 235B A22B Abliterated GGUF?
The files are on Hugging Face in the mradermacher/Qwen3-235B-A22B-abliterated-i1-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 235B A22B 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 Qwen3 235B A22B Abliterated on 8 GB VRAM?
No, the smallest quant is 47.95 GB, which exceeds 8 GB VRAM. You need at least 50 GB VRAM for the smallest available quant.
Does this model support vision tasks?
No, the current variant does not include vision-language capabilities.
What license applies to this model?
The model is released under the Apache‑2.0 license.