Venice Uncensored 1.1

• 🧠 24B parameters, fp16 precision
• 📏 32K token context window
• 📜 Apache 2.0 licensed
• 🔓 Tuned for minimal content restrictions
• 🎨 Built for roleplay and open-ended creative dialogue
• 🌍 Web search capability enabled

Venice

Venice is a privacy-focused AI platform that develops its own in-house models alongside hosting third-party offerings. The organization emphasizes uncensored, unrestricted text generation as a core design principle, building models specifically tuned to…

Venice Uncensored 1.1 is a 24B-parameter text model released in March 2025, purpose-built for creative freedom and unfiltered dialogue. It carries the "most uncensored" trait in the catalogue, making it the go-to option for roleplay, open-ended exploration, and scenarios where typical assistant guardrails get in the way of authentic interaction. The model ships under the permissive Apache 2.0 license at fp16 precision with a 32K token context window and supports web search augmentation.

Within the Venice lineup, this 1.1 release is the prior generation of the Uncensored line, now joined by Venice Uncensored 1.2 (released April 2026). A specialized Venice Role Play Uncensored variant is also available for users who want tuning skewed further toward interactive character work, and an end-to-end encrypted deployment of the same 1.1 weights exists for users prioritizing privacy.

Best suited for writers, roleplayers, and researchers exploring long-form creative scenarios or character-driven dialogue where a compliant, permissive tone matters more than instruction-tuned strictness. Users who want the latest behavior should look to 1.2, but 1.1 remains a solid choice for those who have already tuned prompts around its style.