A GitHub repo rounds up 19 uncensored AI models built for offensive security work
Uncensored and Offensive Security AI Models Benchmark
This curated list catalogs open-weight, uncensored models fine-tuned for authorized red teaming, pentesting, and security research. It covers 19 models ranging from 1.5B to 320B parameters, with details on base models, context lengths, VRAM requirements, uncensoring methods, and training data. Highlights include DeepHat V2, trained on 1.7 million offensive and defensive samples, and pentest-v2, which hits 100% accuracy on GTFOBins versus 25% for its base model.
Curated list of open-weight uncensored models for authorized red team operations, penetration testing, and security research.
- BrawnyBadger53
I can only assume this whole post is meant to be an ad for the cyber frost model? The charts being unreadable such that only cyber frost is identifiable, benchmarks being chosen to mostly support it, and the model being only 2 days old all makes me rather suspect.
- girvo
I’ve been playing with Qwen 3.8 Flash Next uncensored (using the Heretic v2 method) for security exploration, and have been quite impressed, so I’m not surprised to see it near or at the top here. But also it’s a far more powerful base model, so that shouldn’t be too surprising either.
- bede
Please use a categorical colour palette when visualising data like these