G

Gemma-4-31B-Wicked-Stuff-Final

RP Chat

No ratings yet(0)
Prose
—(0)
Intelligence
—(0)
Knowledge
—(0)
Instruction Following
—(0)
Coding
—(0)
Steerability
—(0)

Ratings reflect this model as run on Arli AI.

Gemma-4-31B-Wicked-Stuff-Final is a text generation model available on Arli AI. It offers 31B parameters, 262K token context, reasoning support, vision input. Recommended use: RP Chat. Run it through the OpenAI-compatible chat completions API with unlimited requests and zero-log privacy.

View on Hugging FaceBack to Models

Hourly Usage

Token Usage

Input (24H)

2M

Output (24H)

6K

Input (7D)

18M

Output (7D)

77K

Performance Metrics

Typical Total Time

24.46s

Typical TTFT

3.75s

Typical Input TPS

1498.73

Typical Output TPS

16.18

Model Information

Context Size

262144

Quantization

r128 on FP8

Engine

vllm

Creation Method

Merge

Model Class

Gemma-4-31B - LoRA

Chat Template

Gemma4

Reasoning

Yes

Vision

Yes

Parameters

31B

Added At

10/31/2026

Use Gemma-4-31B-Wicked-Stuff-Final with the Arli AI API

Gemma-4-31B-Wicked-Stuff-Final is available through an OpenAI-compatible chat completions endpoint. Point any OpenAI SDK at https://api.arliai.com/v1, pass Gemma-4-31B-Wicked-Stuff-Final as the model, and requests are served without logging.

cURL

curl https://api.arliai.com/v1/chat/completions \
  -H "Authorization: Bearer $ARLIAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "Gemma-4-31B-Wicked-Stuff-Final",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

Python

from openai import OpenAI

client = OpenAI(
    base_url="https://api.arliai.com/v1",
    api_key="YOUR_ARLIAI_API_KEY",
)

response = client.chat.completions.create(
    model="Gemma-4-31B-Wicked-Stuff-Final",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)

See the text generation docs for every parameter, or read the quick start guide to make your first request.


license: apache-2.0 language:

  • en base_model:
  • ReadyArt/gemma-4-31B-it-scotoma-2 base_model_relation: finetune pipeline_tag: text-generation tags:
  • nsfw
  • explicit
  • roleplay
  • unaligned
  • ERP
  • Erotic
  • Horror
  • Violence
  • adult-content
  • mature
  • conversational
  • instruct

Wicked Stuff 31B

✨ Model Specifications

📚
Dataset
Wicked Tutu + Wicked Stuff
🖤
Focus
Dark
🔄
Epochs
3
⚡
LoRA Rank
128
📝
Prompts
5605
⛓️
Kink List
6,000+

⚙️ Sampler Presets

Temperature
1.0
Top-P
0.92

🚀 Capabilities & Expectations

Wicked Stuff is a specialized instrument for high-intensity roleplay. This is Wicked Tutu (M -> F) mixed in with F -> M prompts bringing balance to the dataset. It is designed to push boundaries and embrace the darker side of storytelling with a strict focus on immersive dynamics.

  • ✅ Unfiltered Explicit Prose
  • ✅ Complex Power Dynamics
  • ✅ Dark Psychological Themes
  • ✅ High-Fidelity Kink Accuracy
  • ✅ Long-Form Narrative Flow
  • ✅ Unaligned & Uncensored (RP)

🛠️ Prompting Guide

Since this is a smut tune and not a general-purpose decensored model, it requires a strong Roleplay (RP) framework to activate its full potential. Use a detailed system prompt to set the scene.

💡 Tip: The more descriptive your initial prompt, the more 'wicked' the output will be.

💡 Usage Tip

For best results, please use Chat Completion and not Text Completion.

🧑‍🔬 The Architects

  • FrenzyBiscuit (Lead/Tuner)
  • GECFDO (Generation)
  • Nimbz (Generation)
  • Kurgan1138 (Prompts)
  • Karki (Template)

🔖 Legal Agreement

  • Accept full responsibility for all generated content.
  • You must be 18+ years old.
  • Architects bear no responsibility for your corruption.
  • Personal use only (non-profit/non-commercial).