Order determines how many previous values to look at when generating the next value. High order will produce an output more similar to the input, while low order will produce a more random result.

Determines how to interpret the source data, which depends on the input format. Lines work best if the input data has a new value on each line, such as the presets, while words and sentences work better for raw text input.

No source data loaded

Load data to begin

Not allowing duplicates will skip over generations that also appear in the source data. This means that each generated word or sentence will be unique, rather than a copy of something in the source material.