GLM-5.3-Flash-CYBERSECURITY
GLMuncensored
from zai-org/GLM-5.3-Flash
↓ 65.2K♥ 464 repos
GLM-5.3-Flash-Uncensored
GLM-5.3-Flash-Uncensored is a abliterated and uncensored variant of zai-org/GLM-5.3-Flash, published on Hugging Face by cekal. It is mit licence. It is distributed across 43 repositories in FP8, GGUF, NVFP4 and MLX formats, totalling 125.8K downloads.
Specification
- Model name
- GLM-5.3-Flash-Uncensored
- Source repository
- cekal/GLM-5.3-Flash-Uncensored
- Base model
- zai-org/GLM-5.3-Flash
- Publisher
- cekal
- Parameters
- Not determinable from name
- VRAM required
- Unknown — no published parameter count
- Hardware class
- Unknown
- Licence
- mit
- Task
- Text generation
- Context window
- 400K tokens (as documented)
- Ablation scope
- Not stated by the publisher
- Formats available
- FP8, GGUF, NVFP4, MLX, EXL3, AWQ, Compressed tensors, Transformers
- First indexed
- 26 Aug 2026
- Last updated
- 27 Oct 2026
Ablation technique
How GLM-5.3-Flash-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
“We introduce GLM-5.3-Flash, the first natively multimodal model in the GLM-5 series. With 320B total parameters and just 18B active parameters, it outperforms GLM-5.2 across benchmarks and real-world workloads at one-tenth the price, while approaching Claude Opus 4.8 on coding and agentic benchmarks.”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download AliceThirty/GLM-5.3-Flash-UNCENSORED-GGUF --local-dir ./glm-5.3-flash-uncensored
llama-cli -m ./glm-5.3-flash-uncensored/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "cekal/GLM-5.3-Flash-Uncensored"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve cekal/GLM-5.3-Flash-Uncensored --trust-remote-codeDownloads and variants (42)
Every published repository of GLM-5.3-Flash-Uncensored, including quantised re-releases by other authors.
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 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 zai-org/GLM-5.3-Flash’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 is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with cekal, and does not independently verify publisher claims. See responsible use.