MindCraft Studio allows you to customize the default behavior of your models. By setting Model Presets, you can ensure that each time you select a specific engine, it is already configured with your preferred parameters or required trigger words.

1. Base Model Defaults

For Base Models, you can define default values for core generation parameters. This is especially useful for models that are optimized for specific settings.

• Default Image Size: Automatically set the aspect ratio and resolution (e.g., 1024x1024).
• Default Steps: Pre-configure the number of refinement iterations (e.g., 8 steps for Z-Image Turbo and 4 steps for FLUX.2[Klein]).

Edit Model Details
Model Name
Z-Image Turbo (4-bit)
Custom Generation Defaults
Default Image Size
Applies automatically when model is selected
Default Steps
Applies automatically when model is selected
Steps8
CancelSave

2. LoRA Trigger Words

LoRAs often require specific "trigger words" to activate their trained style or subject. Instead of typing these manually every time, you can save them directly into the LoRA's metadata within MindCraft Studio.

When a LoRA with saved trigger words is active, they are automatically injected into your prompt during generation, ensuring you always get the intended results without the extra typing.

Edit Model Details
Model Name
Lora Name
Trigger Words
Trigger words...
Trigger words will be automatically added to your prompt.
Model Architecture
Defines model compatibility
FLUX.1
CancelSave

Manual Architecture Adjustment

MindCraft Studio automatically detects the compatible architecture for most LoRAs. However, if the type detection fails or is incorrect, you can manually select the Model Architecture to ensure the LoRA loads correctly with your chosen base model.

You can access these settings by clicking the icon on any model card in the Model Library.

---

Next: Master the professional controls in The Prompt Editor.