TokenPad
Images

Seed and Variation Planner

Change one variable at a time, or learn nothing.

Base prompt

67 characters1 lines0 tokensor drop a file

Seed PlannerExact
0Token change
Token change0no change
Input tokens0what you pasted
Output tokens0what you would send
Token cost of this result
Output tokens0
As input$0.00
× 100K requests$0.00

Everything on this page runs in your browser. Nothing you paste is transmitted, because there is no server here to transmit it to.

Result

The mistake this prevents

Changing the seed and the prompt at the same time is the most common error in image prompt work. The output is different, and you have no way to know which change caused it.

Holding the seed constant while sweeping one variable makes the difference attributable. It is the same discipline as any other experiment.

Why several seeds per value

A single seed can flatter or ruin a value by chance. Three seeds per value gives you enough to see whether a difference is consistent or an artefact of one particular noise pattern.

Run the same seed numbers for every value so the comparison is like for like across the whole sweep.

Frequently asked questions

What is a seed?
The starting random noise the model denoises into an image. The same seed with the same prompt and settings reproduces the same image, which is what makes controlled comparison possible.
How many values should I sweep at once?
Three to five. Beyond that the comparison stops being readable and you are generating more than you will actually look at carefully.

More images tools