Variator Block

How It Works

The Variator acts as a pass-through splitter for your workflow. It accepts one media input (image, video, or 3D) plus optional text context, then creates multiple isolated variation outputs - each passing both the original media and variation-specific text to connected blocks.

What Gets Passed Through

Each variation output sends to connected blocks:

• Original media (if connected) - the image, video, or 3D model from your input
• Variation text - the generated variation prompt or your manual entry

This means downstream Image/Video blocks receive full context: the visual reference AND the variation instructions.

Generation Modes

Auto Mode

AI generates variations based on your instructions. Uses:

• Vision to analyze connected images and understand their content
• Text context from connected Prompt Agent / Prompt blocks as instructions or style guidance
• Your variation prompt to know what to vary (style, color, mood, etc.)

The AI reads all connected inputs, understands what you're working with, and generates N variations of your prompt based on the number of slots you've added.

Manual Mode

Write your own variation text for each slot. Full control over exactly what each variation says.

Features

Context-Aware Generation

Vision analyzes connected images while text context provides instructions - the AI combines both to create relevant variations.

Isolated Outputs

Each variation slot has its own output handle. Connect them to different blocks for parallel workflows, or connect one variation to multiple blocks.

Pass-Through Design

Variator doesn't transform your media - it passes it through alongside variation text, enabling downstream blocks to use both.

Getting Started

1. Add a Variator block to your canvas
2. Connect inputs: one media block (Image/Video/3D) and optionally a Prompt Agent or Prompt block for instructions
3. Add variation slots (click + button)
4. Choose mode: Auto (AI generates) or Manual (you write)
5. For Auto: write what to vary (e.g., "create color variations" or "vary the lighting and mood")
6. Generate variations
7. Connect each variation output to downstream blocks

Pro Tips

1. Connect an Image + Prompt Agent block - Variator uses vision to analyze the image and text as instructions
2. One variation output can connect to multiple Image/Video blocks for batch generation
3. Each variation is isolated - connect different variations to different model settings
4. Watch concurrent generation limits when running all connected blocks