One can only dream

You can run uncensored LLMs locally. They're more accurate and faster when the results aren't filtered.