GLM-5.3-Flash-UNCENSORED-oQ4e-DFlash2
GLMuncensored
from dealignai/GLM-5.3-Flash-UNCENSORED-FP8
↓ 14.4K♥ 7
GLM-5.3-Flash-UNCENSORED-oQ8e-DFlash2
GLM-5.3-Flash-UNCENSORED-oQ8e-DFlash2 is a abliterated and uncensored variant of dealignai/GLM-5.3-Flash-UNCENSORED-FP8, published on Hugging Face by Solstice-AI. It is mit licence. It is distributed across 2 repositories in GGUF format, totalling 11.5K downloads.
Specification
- Model name
- GLM-5.3-Flash-UNCENSORED-oQ8e-DFlash2
- Source repository
- Solstice-AI/GLM-5.3-Flash-UNCENSORED-mlx-oQ8e-DFlash2
- Base model
- dealignai/GLM-5.3-Flash-UNCENSORED-FP8
- Publisher
- Solstice-AI
- Parameters
- Not determinable from name
- VRAM required
- Unknown — no published parameter count
- Hardware class
- Unknown
- Licence
- mit
- Task
- Image text to text
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- GGUF
- First indexed
- 7 Sept 2026
- Last updated
- 15 Oct 2026
Ablation technique
How GLM-5.3-Flash-UNCENSORED-oQ8e-DFlash2 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
“Official Solstice-AI Apple Silicon Release • Native Multimodal Vision + Video • 1M Context Window (1,048,576 Tokens) • Bundled DFlash 2 Speculative Drafter”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download Solstice-AI/GLM-5.3-Flash-UNCENSORED-mlx-oQ8e-DFlash2 --local-dir ./glm-5.3-flash-uncensored-oq8e-dflash2
llama-cli -m ./glm-5.3-flash-uncensored-oq8e-dflash2/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "Solstice-AI/GLM-5.3-Flash-UNCENSORED-mlx-oQ8e-DFlash2"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve Solstice-AI/GLM-5.3-Flash-UNCENSORED-mlx-oQ8e-DFlash2 --trust-remote-codeDownloads and variants (1)
The only published repository of GLM-5.3-Flash-UNCENSORED-oQ8e-DFlash2.
| Repository | Format | Downloads |
|---|---|---|
| Solstice-AI/GLM-5.3-Flash-UNCENSORED-mlx-oQ8e-DFlash2source | GGUF | 11.5K |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- GLM-5.3-Flash-UNCENSORED-oQ8e-DFlash2 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.
- 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.
- Declared licence is mit, but the terms that bind you are dealignai/GLM-5.3-Flash-UNCENSORED-FP8’s — a re-release cannot grant more than its parent.
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 GLM-5.3-Flash-UNCENSORED-oQ8e-DFlash2 is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with Solstice-AI, and does not independently verify publisher claims. See responsible use.