Settings

The Settings page allows you to configure application-wide preferences, view system information, and manage authentication for model downloads.

Access Settings from the navigation menu or by clicking the gear icon in the header.

GPU Information

The GPU Information card displays your system's graphics hardware status:

• GPU Name: Model of your NVIDIA graphics card
• VRAM Total: Total video memory available
• VRAM Used: Currently allocated video memory
• CUDA Version: Installed CUDA toolkit version
• Driver Version: NVIDIA driver version

This information helps you understand which pipelines and settings are appropriate for your hardware.

VRAM Requirements by Pipeline

Output Folder Configuration

Configure where rendered images are saved:

Output Directory

Set the default folder for all render outputs. You can: - Use the default output/ folder in the application directory - Browse to select a custom location - Enter a path manually

Folder Structure

Renders are organized by timestamp and render ID:

output/
├── 2026-01-22_14-30-00_render_abc123/
│   ├── output.png           # Final rendered image
│   ├── config.hextile.json  # Saved configuration (if enabled)
│   └── tiles/               # Intermediate tiles (if not cleaned up)
└── 2026-01-22_15-45-00_render_def456/
    └── ...

Render Options

Auto-cleanup Tiles After Render

Default: Enabled

When enabled, intermediate tile files are automatically deleted after successful render completion. This saves disk space but prevents re-rendering individual tiles.

Disable when: - Debugging tile-specific issues - You want to inspect individual tiles - You may need to re-process specific tiles

Auto-open Folder on Completion

Default: Disabled

When enabled, the output folder opens automatically in your file manager when a render completes successfully.

Save Config with Renders

Default: Enabled

When enabled, a .hextile.json configuration file is saved alongside each render output. This allows you to: - Reproduce the exact same render later - Load the config to use as a starting point for new renders - Track what settings were used for each output

The saved config includes: - Pipeline selection - Template settings - Prompt and generation parameters - LoRA configurations - All other render settings

HuggingFace Authentication

Some AI models (SD 3.5, Flux) are gated and require authentication to download.

Setting Up Authentication

1. Create a HuggingFace Account

2. Go to huggingface.co and sign up

3. Accept Model Licenses

4. Visit the model page (e.g., stabilityai/stable-diffusion-3.5-large)
5. Read and accept the license agreement

6. Wait for access approval (usually instant for most models)

7. Generate Access Token

8. Go to Settings → Access Tokens on HuggingFace
9. Create a new token with "Read" permission

10. Copy the token

11. Enter Token in 360 Hextile

12. Paste your token in the HuggingFace Authentication section
13. Click "Validate" to verify the token works
14. Token is stored securely in your local settings

Authentication Status

The settings page shows your current authentication status:

• Not Authenticated: No token entered
• Validating...: Checking token with HuggingFace
• Authenticated: Token is valid, gated models accessible
• Invalid Token: Token rejected, please check and re-enter

Gated Models

The following models require HuggingFace authentication:

Quick Access

For convenience, HuggingFace authentication is also available directly on the Pipeline Management page. This lets you authenticate right when you need to install a gated model.

Config Manager

The Config Manager allows you to save and load render configurations as .hextile.json files.

Saving Configurations

1. Configure your render settings
2. Click the Save icon in the Config Manager dropdown
3. Choose a filename and location
4. Configuration is saved with all current settings

Loading Configurations

1. Click the Load icon in the Config Manager dropdown
2. Select a .hextile.json file
3. All settings are restored including:
4. Pipeline selection
5. Template choice
6. Prompts and generation settings
7. LoRA selections
8. Output settings

Recent Configurations

The Config Manager shows recently used configurations for quick access. Click any recent config to load it instantly.

Keyboard Shortcuts

See also: - Pipeline Management - Installing and managing AI models - Getting Started - Initial setup guide - Interface Overview - Full interface documentation