Venice Role Play Uncensored
Venice's fine-tuned 24B roleplay model, optimized for highly expressive, low-refusal character interactions with zero retention.
Overview
What is Venice Role Play Uncensored
Venice Role Play Uncensored is a 24B-parameter creative model released in February 2026. Built on an open-weights architecture, it is specifically fine-tuned for immersive, highly expressive, and character-driven roleplay, minimizing the rigid refusals and safety-nanny filters common in mainstream Big Tech models.
Running it privately on Venice
Running Venice Role Play Uncensored on Venice ensures absolute sovereignty over your creative writing. Your prompts are processed with zero retention, meaning your private fantasies, stories, and character interactions are never logged, analyzed, or used to train future models. It is a completely permissionless creative sandbox.
Assessment
Strengths and limitations
- Highly expressive and immersive character-driven dialogue with rich emotional nuance.
- Drastically minimized refusals, allowing for creative freedom and complex narrative themes without Big Tech censorship.
- Equipped with Vision capabilities, allowing you to feed character portraits or scene references directly into the roleplay.
- Supports tool use, web search, and structured JSON outputs for building advanced interactive text games or agents.
- Generous 128K context window to maintain deep memory of long-running roleplay scenarios.
- At 24B parameters, it may occasionally struggle with highly complex logic or advanced coding compared to massive frontier models.
- Not hosted in a Trusted Execution Environment (TEE) or end-to-end encrypted, though it still benefits from Venice's strict zero-retention policy.
- Optimized primarily for creative writing and roleplay, making it less ideal for dry, academic, or highly structured enterprise tasks.
Capabilities
What it supports
- Tool use / function calling
- Vision (image input)
- Reasoning
- Web search
- Code-optimized
- Structured output (JSON schema)
- Audio input
- Video input
- Multiple image inputs
- Log probabilities
Specifications
Datasheet
- Maker
- Venice AI
- Released
- February 20, 2026
- Architecture
- Fine-tuned 24B Parameter Model
- Base Model
- dphnAI 24B-3.2-RP-K2-final
- License
- Apache-2.0
- Open weights
- Yes
- Quantization
- FP8
- Context window
- 128K tokens
- Max output
- 4.096K tokens
- Capabilities
- Vision, Function calling, Web search
- Privacy on Venice
- Private — zero retention
- Available on Venice since
- Feb 2026
API
Call it from your code
Venice exposes an OpenAI-compatible API. Point your base URL at Venice and pass the model id.
curl https://api.venice.ai/api/v1/chat/completions \
-H "Authorization: Bearer $VENICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "venice-uncensored-role-play",
"messages": [{ "role": "user", "content": "Explain quantum tunneling simply." }]
}'Pricing
What it costs on Venice
Billed per token on Venice: $0.50 per 1M input tokens and $2 per 1M output tokens.
New Venice accounts include a free daily allowance and 500 welcome credits — no credit card required.
Alternatives
How it compares
| Model | Best for | Context window | Open weights | Vision support | Price (Venice) |
|---|---|---|---|---|---|
| Venice Role Play Uncensored | Purpose-built for expressive, low-refusal roleplay with vision. | 128K tokens | Yes | Yes | $0.50 in · $2 out / 1M |
| DeepSeek V3.2 | Incredibly cheap open-weights alternative, but less specialized for creative roleplay. | 160K tokens | Yes | Yes | $0.33 in · $0.48 out / 1M |
| Google Gemma 4 31B Instruct | Larger open-weights model with massive context, but carries standard alignment filters. | 256K tokens | Yes | Yes | $0.12 in · $0.36 out / 1M |
| Claude Sonnet 4.6 | Massive context and high intelligence, but highly censored and expensive. | 1M tokens | No | Yes | $3.60 in · $18 out / 1M |
Purpose-built for expressive, low-refusal roleplay with vision.
Use cases
What it is good for
- 01Immersive, multi-character interactive fiction and text-based roleplay.
- 02Creative writing, brainstorming, and drafting uncensored dialogue for novels or screenplays.
- 03Building custom interactive text adventure games utilizing structured JSON outputs.
- 04Visual roleplay using character sheets, portraits, or scene artwork via its vision capabilities.
Prompting
Getting better results
Use detailed system prompts to define your character's personality, tone, and background before starting the chat.
Provide visual references (like character portraits) to ground the model's descriptions of physical appearance.
If the model drifts, use the 128K context window to occasionally summarize the story so far and keep the narrative on track.
Version history
General-purpose 24B uncensored model.
Current — roleplay-tuned 24B model with vision and tools.
FAQ
Frequently asked questions
Venice Role Play Uncensored is a 24B-parameter open-weights model fine-tuned by Venice AI (based on dphnAI's 24B-3.2-RP-K2-final). It is optimized specifically for creative writing, storytelling, and immersive character roleplay.
It is billed per token on Venice at $0.50 per 1M input tokens and $2.00 per 1M output tokens.
The model is built on open weights (licensed under Apache-2.0). Venice offers a free tier with daily prompt limits, while unlimited or high-priority access is available via Venice Pro or pay-per-token API pricing.
Yes, it supports vision capabilities, including multiple image inputs. You can upload character portraits, scene references, or setting art to enhance your roleplay experience.
Venice operates under a strict zero-retention privacy policy. Your prompts, character definitions, and private roleplay chats are never stored on servers, profiled, or used to train models, giving you complete creative sovereignty.
Claude Sonnet 4.6 has a larger context window and higher general intelligence, but it is heavily censored, proprietary, and expensive. Venice Role Play Uncensored is open-weights, highly affordable, and designed to minimize refusals, making it far superior for creative, unrestricted roleplay.
Yes, Venice Role Play Uncensored supports tool use (function calling), structured JSON output, and web search, allowing it to act as an interactive game master or fetch real-world information.
Run Venice Role Play Uncensored privately
No prompt logging. No data used for training.