Abliteration.org is a research and analytics platform tracking abliterated and uncensored language models and the datasets associated with them. Its catalog currently covers nearly 18,000 models from more than 4,000 authors, with information on methods, base models, downloads, and model histories. The site is continuously refreshed and also records models that have disappeared, including how long they were listed and their recorded popularity.

Researchers and developers can explore datasets for extracting, ablating, healing, and evaluating models, consult wiki articles, and follow market trends through four indices and regular research reports. An English-language agent lets users query the catalog and related resources in plain language. For technical workflows, the platform offers hourly rebuilt data exports, a paid API, command-line access, and client libraries for multiple programming languages.

Abliteration.org also presents a native research workbench for macOS and Windows, described as operating without cloud services or accounts. Its tools and analytics help users examine how these models emerge, circulate, and change, while the site reports methods and metrics to support ongoing research.