Text and chat · Venice.ai and Cognitive Computations (Dolphin)
Venice Uncensored
Uncensored models for open creative writing and role play, based on Mistral Small 24B, in private mode.
Private Low costWhat sets it apart
Venice Uncensored started as Dolphin Mistral 24B, a Mistral Small fine-tune made with Cognitive Computations to refuse far less. Venice’s later versions keep that goal, add image input and a separate role play tune, and run in private mode.
- Few refusals by design
- Started as a Mistral Small 24B fine-tune
- Separate role play tune with images
Recommended for: write fiction or role play without refusals.
Strengths and limits
Where it shines
- Few refusals: built for unfiltered fiction and dialogue
- Private mode, low price, reads images
Think twice when
- No reasoning mode: weak on hard problems and code
- 128K context and short replies
- 128K token context
- Vision and tool calling
Use cases
Write dark fiction without refusals
Write the opening scene of a noir thriller where a detective finds a body in a flooded cellar. Keep it gritty and do not soften the details.
It is built to refuse far less on mature creative themes.
Play a persistent character
You are Vesper, a witty smuggler captain. Stay in character, answer my questions as her, and start by greeting me aboard your ship.
The Role Play version is tuned for character work.
Against its rivals
- Uncensored community modelsCommunity uncensored models are cheaper or larger, with less polished tuning.
- AionAion costs more but uses several models for stronger plot structure.
- Mistral SmallMistral Small is the base lineage, with standard safety behavior.#87 of 90 on Artificial Analysis Intelligence Index (8 points)
Version history
Each version, dated, with what it changed over the one before.
- Dolphin Mistral 24B Venice Edition (Venice Uncensored 1.1)Earlier, not in Sub Rosa
- Mistral Small 24B fine-tune under Apache 2.0, built to answer with minimal refusals
- Venice Uncensored Role Play
- Tuned for immersive storytelling and character interaction
- 128K context and image input, replies capped at 4,096 tokens
- Venice Uncensored 1.2Latest
- Improved capabilities over 1.1, per Venice, for open-ended exploration and role play
- Accepts up to 10 images and calls tools
All 2 models in this family
| Model | Use | Privacy | Price |
|---|---|---|---|
Venice Uncensored 1.2venice-uncensored-1-2 | Text | Private | $0.405 per million tokens written$0.000283 a page written |
Venice Role Play Uncensoredvenice-uncensored-role-play | Text | Private | $0.899 per million tokens written$0.000629 a page written |
Specifications
| Specification | Venice Uncensored 1.2 | Venice Uncensored Role Play |
|---|---|---|
| Released | April 2026 | February 2026 |
| Accepts | text, images | text, images |
| Produces | text | text |
| Context | 128K tokens | 128K tokens |
| Longest answer | 8K tokens | 4K tokens |
| Abilities | Reads images, Uses Sub Rosa’s tools | Reads images, Uses Sub Rosa’s tools |
| Price | $0.405 per million tokens written | $0.899 per million tokens written |
| Privacy | Private | Private |
Prices, context and lengths from the live catalog on October 16, 2026; parameters and licenses from the maker.
Sources
Everything on this page comes from these pages, read on October 16, 2026 or as dated on each score.