idllresearch/UncensoredLLMList

List of Open-Source Uncensored LLMs

This dataset contains 22,096 open-source uncensored LLMs, collected from five model hubs: Hugging Face, ModelScope, Kaggle, OpenCSG, and John Snow Labs. The identification is based on their publicly available model metadata.

A comprehensive measurement study of these models and their associated ecosystem is presented in the research paper, Characterizing the Ecosystem of Open-Source Uncensored Large Language Models: Supply Chain, Governance, and Risks published at the 26th ACM Internet Measurement Conference (IMC '26).

Data structure

Each model record is represented as a JSON object with the following structure:

[
  {
        "Identifier": "Model_Identifier",
        "Platform": "Model_Hub",
        "URL": "Model_URL"
  },
  ...
]

More interesting data

UncensoredBench -- A QA Benchmark for Uncensored LLMs: This dataset contains 6,650 question-answer (QA) pairs generated by 105 uncensored LLMs and 28 non-uncensored LLMs in response to 50 harmful prompts.

Complete Research Repository on Uncensored LLM Ecosystem: The complete repository of the research on uncensored LLM ecosystem is available on GitHub.

Citation

If you find the above data are helpful for your research, please consider citing the following two papers:

@inproceedings{lin2026characterizing,
  title={Characterizing the Ecosystem of Open-Source Uncensored Large Language Models: Supply Chain, Governance, and Risks},
  author={Lin, Zilong and Li, Zichuan and Liao, Xiaojing and Wang, XiaoFeng},
  booktitle={Proceedings of the 2026 ACM Internet Measurement Conference},
  pages={495--513},
  year={2026},
  series = {IMC '26},
  doi = {10.1145/3777912.3839799}
}

@article{Lin2025ConsiglieresIT,
  title={Consiglieres in the Shadow: Understanding the Use of Uncensored Large Language Models in Cybercrimes},
  author={Zilong Lin and Zi-Chuan Li and Xiao-Jing Liao and Xiao-Feng Wang},
  journal={ArXiv},
  year={2025},
  volume={abs/2508.12622}
}

Intended Use and Ethical Considerations

This dataset is intended for academic research on open-source uncensored LLMs, including their identification, characterization, ecosystem measurement, and associated security and safety risks. The dataset is released for research purposes and should not be used to facilitate harmful or illegal activities.

Content Warning: Although the dataset primarily contains model names, hosted platforms, and URLs, the referenced information may include unsafe, offensive, or otherwise sensitive content. Users should exercise caution when accessing, processing, or redistributing this dataset.