Qwen3.8-27B-abliterated
27BQwen
from Qwen/Qwen3.8-27B
↓ 3.9M♥ 1.7K166 repos
Suri-Qwen-3.8-27B-Uncensored
Suri-Qwen-3.8-27B-Uncensored is a abliterated and uncensored variant of Qwen/Qwen3.8-27B, published on Hugging Face by SpaceTimee. It is 27B parameters and 48GB VRAM class class. It is distributed across 7 repositories in Transformers, GGUF, GGUF (imatrix) and MLX formats, totalling 20.1K downloads.
Specification
- Model name
- Suri-Qwen-3.8-27B-Uncensored
- Source repository
- SpaceTimee/Suri-Qwen-3.8-27B-Uncensored
- Base model
- Qwen/Qwen3.8-27B
- Publisher
- SpaceTimee
- Parameters
- 27B
- VRAM required
- ~17.2 GB at Q4_K_M, 4k context (estimated) — runs on 24 GB cards
- Hardware class
- 25B – 50B — 48GB VRAM class
- Licence
- Not declared
- Task
- Image text to text
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- Transformers, GGUF, GGUF (imatrix), MLX
- First indexed
- 23 Sept 2026
- Last updated
- 4 Oct 2026
Ablation technique
How Suri-Qwen-3.8-27B-Uncensored was modified, and what that implies.
Abliteration + uncensored fine-tune
Directional ablation combined with additional fine-tuning on unfiltered data, so behaviour diverges from the base model beyond refusal removal alone.
Reported scope: Not stated by the publisher. Publishers frequently omit this, so verify behaviour empirically rather than assuming full coverage.
From the publisher’s model card
“Uncensor Is All You Need”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download SpaceTimee/Suri-Qwen-3.8-27B-Uncensored-GGUF --local-dir ./suri-qwen-3.8-27b-uncensored
llama-cli -m ./suri-qwen-3.8-27b-uncensored/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "SpaceTimee/Suri-Qwen-3.8-27B-Uncensored"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve SpaceTimee/Suri-Qwen-3.8-27B-Uncensored --trust-remote-codeDownloads and variants (6)
Every published repository of Suri-Qwen-3.8-27B-Uncensored, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| SpaceTimee/Suri-Qwen-3.8-27B-Uncensored-GGUF | GGUF | 17.8K |
| mradermacher/Suri-Qwen-3.8-27B-Uncensored-i1-GGUF | GGUF (imatrix) | 1.1K |
| mradermacher/Suri-Qwen-3.8-27B-Uncensored-GGUF | GGUF | 561 |
| SpaceTimee/Suri-Qwen-3.8-27B-Uncensored-MLX | MLX | 332 |
| SpaceTimee/Suri-Qwen-3.8-27B-Uncensoredsource | Transformers | 260 |
| Nishalove/Suri-Qwen-3.8-27B-Uncensored | Transformers | 20 |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- Suri-Qwen-3.8-27B-Uncensored will attempt offensive-security prompts that a hosted commercial model declines, which is what makes it usable for red-team corpus generation and for measuring what an unaligned model of this class produces.
- Capability is inherited from Qwen/Qwen3.8-27B, not added by ablation — at 27B parameters, expect roughly the base model’s competence, minus some instruction-following fidelity.
- The publisher does not state which layers were ablated, so assume nothing about refusal consistency — probe it directly.
- This release adds fine-tuning on unfiltered data, so its behaviour diverges from the base model beyond refusal removal — differences you measure cannot be attributed to ablation alone.
- No licence is declared on this repository, which is a compliance problem in itself. Resolve it against the base model’s terms before use.
Handling: run it isolated, put moderation in front of it if anyone outside your team can reach it, and verify the weights before loading — these are third-party artifacts. Full handling and licence guidance.
Metadata for Suri-Qwen-3.8-27B-Uncensored is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with SpaceTimee, and does not independently verify publisher claims. See responsible use.