Ganbot Commands

Ganbot is a multi-platform bot with AI-powered image generation capabilities.

Image Generation Commands

!prompt - Direct Image Generation

Generate images with precise control over model and parameters.

Basic Syntax:

!prompt <description> [options]

Available Options:

• -m <model> or --model <model> - Select the model to use (e.g., flux, sdxl)
• -c <number> or --count <number> - Number of images to generate (default: 1)
• -w <pixels> or --width <pixels> - Image width
• -h <pixels> or --height <pixels> - Image height
• --ar <ratio> or --aspect <ratio> - Aspect ratio (e.g., 16:9, 1:1, 9:16)
• --seed <number> - Random seed for reproducibility
• -s <number> or --steps <number> - Number of inference steps
• --denoise <0.0-1.0> - Denoising strength for img2img
• --alias <name> - Load settings from a saved alias
• --no <terms> - Negative prompt (everything after --no is excluded from the image)

Advanced Options:

• --sampler <name> - Sampler algorithm to use (e.g., euler, dpmpp_2m, er_sde)
• --scheduler <name> - Scheduler to use (e.g., normal, karras, beta, sgm_uniform)

Option Shortcuts:

You can combine short options with their values:

• -w512 instead of -w 512
• -h768 instead of -h 768
• -mflux instead of -m flux

You can also use equals signs with long options:

• --width=1920 instead of --width 1920
• --model=sdxl instead of --model sdxl

Examples:

Simple generation:

!prompt a majestic dragon flying over snow-capped mountains

Specify a model:

!prompt a cyberpunk city at night -m flux

Control dimensions with aspect ratio:

!prompt beautiful landscape --ar 16:9 -m sdxl

Control dimensions precisely:

!prompt portrait of a wizard -w 512 -h 768

Generate multiple images at once:

!prompt cute robot -c 4 -m flux

Use negative prompts to exclude elements:

!prompt beach scene --no umbrella people crowds

Full control with multiple options:

!prompt futuristic spaceship -m flux --ar 16:9 -s 50 --seed 42 --no blur text

Reproducible generation:

!prompt sunset over ocean --seed 12345 -m sdxl

!dream - Artistic Variations

Generate multiple artistic variations of your concept with creative AI-enhanced prompts and detective-themed commentary.

Basic Syntax:

!dream <description> [options]

How it Works:

!dream takes your request and creates multiple unique variations by:

1. Generating distinct, detailed prompts for each variation
2. Creating images from those enhanced prompts
3. Presenting them in a gallery with creative commentary

Key Differences from !prompt:

• Defaults to 2 variations (controlled with -c, max 6)
• Uses the "dream" model alias by default
• Enhances your prompt with thematic details and atmosphere
• Includes creative commentary with results
• Only works with English-friendly models

Supported Options:

All !prompt options work, but the most useful are:

• -m <model> - Select a different model
• -c <number> - Number of variations (1-6, default: 2)
• --ar <ratio> - Aspect ratio
• -w/-h - Dimensions

Examples:

Basic dream request (generates 2 variations):

!dream a mysterious forest at twilight

More variations:

!dream steampunk airship -c 4

Different model and aspect ratio:

!dream neon-lit alleyway -m flux --ar 16:9

Landscape with specific dimensions:

!dream ancient ruins -c 3 -w 1024 -h 768

Getting Help

For more information, visit the Model Gallery to see examples of what different models can generate.