🇬🇧 EN

🎲 Random

Generates a random integer within a range, or selects a random item from a list. Useful for A/B testing, random routing, sampling, and load distribution.

Category: Flow Control · Type identifier: random

Overview

The Random node introduces controlled non-determinism into your workflows. You can use it in two modes:

Number mode — generates a random integer between a minimum and maximum value (both inclusive). Great for generating reference numbers, test data, or probability thresholds.
Choice mode — picks one item at random from a list you define. Each item in the list has an equal chance of being selected. Ideal for A/B or multivariate testing, random message templates, or distributing work across multiple targets.

The result is available as a variable and can be passed to a Switch node to route execution based on which variant was selected, or used directly in a template.

Configuration

Field	Status	Description
Mode	Required	`number` — generate a random integer; `choice` — pick one item from a defined list.
Min	Conditional	Required when Mode is `number`. The minimum value (inclusive). Supports `{{ variable }}` references.
Max	Conditional	Required when Mode is `number`. The maximum value (inclusive). Must be greater than or equal to Min. Supports `{{ variable }}` references.
Choices	Conditional	Required when Mode is `choice`. An array of values to select from. Each entry can be a static string, number, or a `{{ variable }}` reference. All items have equal probability.
Seed	Optional	A fixed integer seed for the random number generator. When set, the same seed always produces the same result. Useful for testing workflows reproducibly without changing the node configuration.

Output Data

Variable	Type	Description
`value`	integer or string	The generated number (in number mode) or the selected item (in choice mode).
`mode`	string	The mode that was used: `"number"` or `"choice"`.
`index`	integer	In choice mode: the zero-based position of the selected item in the Choices list. Always `null` in number mode.

// Access random outputs (replace "random_step" with your step key) {{ random_step.output.value }} {{ random_step.output.mode }} {{ random_step.output.index }}

Example Usage

A/B test — route users to one of two variants

Add a Random node (step key: ab_picker). Set Mode to choice and add two choices: variant_a and variant_b.
Add a Switch node after it. Set the Expression to:
{{ ab_picker.output.value }}
Add cases variant_a and variant_b.
Connect each Switch branch to the appropriate email template or action node. Each user run will randomly receive one variant.

Generate a random verification code

Mode: number Min: 100000 Max: 999999 // Use the result in a Send Email node Your verification code is: {{ verify_code.output.value }}

Tips & Notes

Equal probability only. All choices have the same chance of being selected. If you need weighted randomness (e.g. 70% variant A, 30% variant B), use a Number mode with a range of 1–10 and a Logic Gate — route numbers 1–7 to variant A and 8–10 to variant B.
Seed for reproducible testing. Set a Seed value during development to get consistent results across test runs. Remove or change the seed in production to restore true randomness.
Combine with Switch. The Random node on its own just picks a value. To actually route execution based on that value, pipe the output into a Switch node — this is the standard pattern for random branching.

Related Nodes

Switch — route execution based on the random value produced.
Logic Gate — apply threshold checks to a random number (e.g. value < 5 for weighted probability).