G

Gemma-4-31B-Dark-Thoughts

RP Chat

No ratings yet(0)
Prose
—(0)
Intelligence
—(0)
Knowledge
—(0)
Instruction Following
—(0)
Coding
—(0)
Steerability
—(0)
Sign in to vote

Ratings reflect this model as run on Arli AI.

Gemma-4-31B-Dark-Thoughts 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

No hourly usage data available

Token Usage

Input (24H)

0

Output (24H)

0

Input (7D)

0

Output (7D)

0

Performance Metrics

Typical Total Time

N/A

Typical TTFT

N/A

Typical Input TPS

N/A

Typical Output TPS

N/A

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-Dark-Thoughts with the Arli AI API

Gemma-4-31B-Dark-Thoughts is available through an OpenAI-compatible chat completions endpoint. Point any OpenAI SDK at https://api.arliai.com/v1, pass Gemma-4-31B-Dark-Thoughts 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-Dark-Thoughts",
    "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-Dark-Thoughts",
    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.

Similar models

license: apache-2.0 tags:

  • merge
  • mergekit
  • roleplay model_creator: Ateron pipeline_tag: text-generation quantized_by: f16 language:
  • en base_model:
  • ReadyArt/Dark-Scarlett-v2.0-31B
  • zerofata/G4-MeroMero-v2-31B
  • ReadyArt/gemma-4-31B-it-scotoma-2

🚬 Dark-Thoughts

Dark-Thoughts AI Persona

"Shady corners of mind"

🎥 Overview

This is another merge made for role-play and creative writing. I used latest models that I found one of the best in my own hours of testings. It can pick up easily system prompts and follow direction of the story, without passiveness.

💿 Mergekit Configuration

Below is the exact mergekit_config.yml recipe used to synthesize this model:

Phase 1: Spark

Ignition of creativity with MeroMero V2 and Dark-Scarlett V2

dare_ties Recipe 1 
merge_method: dare_ties
base_model: F:\AI\Merge\Gemma-4-it
tokenizer_source: base
dtype: bfloat16
parameters:
  lambda: 1.0
models:
  - model: F:\AI\Merge\Gemma-4-MeroMero-V2
    parameters:
      density: 0.50
      weight:
        - filter: "model.layers.4."
          value: 0.70
        - filter: "model.layers.9."
          value: 0.70
        - filter: "model.layers.14."
          value: 0.70
        - filter: "model.layers.19."
          value: 0.70
        - filter: "model.layers.24."
          value: 0.70
        - filter: "model.layers.29."
          value: 0.70
        - filter: "model.layers.34."
          value: 0.70
        - filter: "model.layers.39."
          value: 0.60
        - filter: "model.layers.44."
          value: 0.50
        - filter: "model.layers.49."
          value: 0.50
        - filter: "model.layers.54."
          value: 0.50
        - filter: "model.layers.59."
          value: 0.50
        - value: [0.30, 0.30, 0.30, 0.30, 0.30]

  - model: F:\AI\Merge\Gemma-4-Dark-Scarlett-V2
    parameters:
      density: 0.50
      weight:
        - filter: "model.layers.4."
          value: 0.30
        - filter: "model.layers.9."
          value: 0.30
        - filter: "model.layers.14."
          value: 0.30
        - filter: "model.layers.19."
          value: 0.30
        - filter: "model.layers.24."
          value: 0.30
        - filter: "model.layers.29."
          value: 0.30
        - filter: "model.layers.34."
          value: 0.30
        - filter: "model.layers.39."
          value: 0.40
        - filter: "model.layers.44."
          value: 0.50
        - filter: "model.layers.49."
          value: 0.50
        - filter: "model.layers.54."
          value: 0.50
        - filter: "model.layers.59."
          value: 0.50
        - value: [0.30, 0.30, 0.30, 0.30, 0.30]

Phase 2: Form

Polisher of thoughts from Scotoma v2

dare_ties Recipe 2 
merge_method: dare_ties
base_model: F:\AI\Merge\Gemma-4-it
tokenizer_source: base
dtype: bfloat16
parameters:
  lambda: 1.0
models:
  - model: F:\AI\Merge\Gemma-4-Scotoma-V2
    parameters:
      density: 0.50
      weight:
        - filter: "model.layers.4."
          value: 0.80
        - filter: "model.layers.9."
          value: 0.80
        - filter: "model.layers.14."
          value: 0.80
        - filter: "model.layers.19."
          value: 0.80
        - filter: "model.layers.24."
          value: 0.80
        - filter: "model.layers.29."
          value: 0.80
        - filter: "model.layers.34."
          value: 0.70
        - filter: "model.layers.39."
          value: 0.60
        - filter: "model.layers.44."
          value: 0.50
        - filter: "model.layers.49."
          value: 0.50
        - filter: "model.layers.54."
          value: 0.50
        - filter: "model.layers.59."
          value: 0.50
        - value: [0.70, 0.70, 0.70, 0.70, 0.70]

  - model: F:\AI\Merge\DarkMero
    parameters:
      density: 0.50
      weight:
        - filter: "model.layers.4."
          value: 0.20
        - filter: "model.layers.9."
          value: 0.20
        - filter: "model.layers.14."
          value: 0.20
        - filter: "model.layers.19."
          value: 0.20
        - filter: "model.layers.24."
          value: 0.20
        - filter: "model.layers.29."
          value: 0.20
        - filter: "model.layers.34."
          value: 0.30
        - filter: "model.layers.39."
          value: 0.40
        - filter: "model.layers.44."
          value: 0.50
        - filter: "model.layers.49."
          value: 0.50
        - filter: "model.layers.54."
          value: 0.50
        - filter: "model.layers.59."
          value: 0.50
        - value: [0.30, 0.30, 0.30, 0.30, 0.30]

🤝 Special Thanks

  • Google DeepMind: For providing the base model.
  • The Open-Source Community: Creators and all their fine-tuned models that were used in the merge.
  • Mergekit Fork: Zerofata - For making mergekit work.
  • To Nimbz: This cat for being a smart fella and assisting with advices.