What is Venice Uncensored?
Venice AI's house uncensored model — the production continuation of Dolphin Mistral 24B Venice Edition, the fine-tune Cognitive Computations built with Venice. Version 1.2 is served at fp16 precision with a 128K context window. Its defining property is trained-in directness: it follows the system prompt absolutely and answers legal requests without the hedging, disclaimers or refusals that mainstream assistants attach, which is why it became the default recommendation whenever someone asks for a genuinely uncensored model.