Instructions to use TaterTotterson/Gemma-4-26B-A4B-IT-UD-Q4_K_XL-mlx-Tater-NoThink with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use TaterTotterson/Gemma-4-26B-A4B-IT-UD-Q4_K_XL-mlx-Tater-NoThink with MLX:

# Make sure mlx-lm is installed
# pip install --upgrade mlx-lm

# Generate text with mlx-lm
from mlx_lm import load, generate

model, tokenizer = load("TaterTotterson/Gemma-4-26B-A4B-IT-UD-Q4_K_XL-mlx-Tater-NoThink")

prompt = "Write a story about Einstein"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True
)

text = generate(model, tokenizer, prompt=prompt, verbose=True)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

How to use TaterTotterson/Gemma-4-26B-A4B-IT-UD-Q4_K_XL-mlx-Tater-NoThink with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "TaterTotterson/Gemma-4-26B-A4B-IT-UD-Q4_K_XL-mlx-Tater-NoThink"

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "TaterTotterson/Gemma-4-26B-A4B-IT-UD-Q4_K_XL-mlx-Tater-NoThink"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use TaterTotterson/Gemma-4-26B-A4B-IT-UD-Q4_K_XL-mlx-Tater-NoThink with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "TaterTotterson/Gemma-4-26B-A4B-IT-UD-Q4_K_XL-mlx-Tater-NoThink"

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default TaterTotterson/Gemma-4-26B-A4B-IT-UD-Q4_K_XL-mlx-Tater-NoThink

Run Hermes

hermes

MLX LM

How to use TaterTotterson/Gemma-4-26B-A4B-IT-UD-Q4_K_XL-mlx-Tater-NoThink with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "TaterTotterson/Gemma-4-26B-A4B-IT-UD-Q4_K_XL-mlx-Tater-NoThink"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "TaterTotterson/Gemma-4-26B-A4B-IT-UD-Q4_K_XL-mlx-Tater-NoThink"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "TaterTotterson/Gemma-4-26B-A4B-IT-UD-Q4_K_XL-mlx-Tater-NoThink",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Gemma-4-26B-A4B-IT UD-Q4_K_XL MLX - Tater NoThink

This is a Tater NoThink repack of Brooooooklyn/Gemma-4-26B-A4B-IT-UD-Q4_K_XL-mlx.

What changed

The model weights are unchanged.
chat_template.jinja now forces enable_thinking = false.
The model card is tagged for Tater Picks auto-discovery.

Why

Tater works best with models that answer directly and do not emit hidden reasoning or thinking blocks. This repo keeps the same MLX quantized model but makes the default chat template run in NoThink mode for runtimes that use the embedded template.

Source

Source model: Brooooooklyn/Gemma-4-26B-A4B-IT-UD-Q4_K_XL-mlx
Base model: google/gemma-4-26b-a4b-it
Format: MLX SafeTensors, UD-Q4_K_XL

License

This model inherits the Gemma Terms of Use from the base model.

Downloads last month: -

Safetensors

Model size

5B params

Tensor type

BF16

U32

MLX

Hardware compatibility

4-bit