Seed Range

Name	Type	Description
バリアント一覧 variants	List<VariantRecord>	生成した variant record 一覧です
スナップショット一覧 snapshots	List<VariantSnapshot>	生成した variant snapshot 一覧です
レシピ recipe	VariantRecipe	共通の生成レシピです
seed一覧 seeds	List<Int>	生成した seed 値一覧です
個数 count	Int	生成した variant 数です

Parameters

Name	Type	Default	Range	Description / Options
キー key	Text	seed	—	snapshot に書き込むキー名です
開始seed start	Int	0	-4096–4096	開始 seed 値です
個数 count	Int	8	1–256	生成する seed の数です
増分 step	Int	1	-1024–1024	seed を増やす増分です

Overview

Batch-produces a list of seed candidates for random nodes. It streamlines rolling through random generations—noise, scattering, generated characters, and more—to compare results with identical settings except for the seed and find a lucky outcome.

Usage tips

The go-to flow is seed_range (N candidates) → variant_select (switch by index). Step through seed variations, compare the results, and keep the winning seed.
For batch production, pass it to batch_render. You can build multi-axis searches such as 4 seeds × 5 strength levels by multiplying the axes with cartesian_product.
A larger step, such as 100, makes differences more likely even in random implementations where adjacent seeds produce similar results.

Related nodes

variant_range — sibling node for numeric parameters
variant_select — select and inspect one candidate
noise_generate and other random nodes — destinations for seeds
batch_render — render the seed list in a batch

Read this first

If VariantRange creates ordinary numeric candidates, SeedRange is a dedicated seed-candidate generator.
This node also creates multiple candidates at once, not just one.

Understanding the outputs

Output	Contents	Use
`seeds`	Raw integer seeds	When you only need the seed numbers
`snapshots`	Application-ready values such as `seed=10`	Pass to `BatchRender` or downstream nodes
`variants`	Saved form with names and recipes attached to snapshots	Save, select, or display as a list
`recipe`	How the entire candidate set was made	Preserve how it was generated

Start with this example

Use these settings:

key = seed
start = 10
count = 4
step = 2

The generated seeds are:

[10, 12, 14, 16]

Seed list

[10, 12, 14, 16]

Snapshot list

[
  { "version": 1, "values": { "seed": 10 } },
  { "version": 1, "values": { "seed": 12 } },
  { "version": 1, "values": { "seed": 14 } },
  { "version": 1, "values": { "seed": 16 } }
]

Variant list

The first entry looks roughly like this:

{
  "version": 1,
  "id": "seed:seed:0",
  "name": "seed=10",
  "favorite": false,
  "snapshot": {
    "version": 1,
    "values": { "seed": 10 }
  },
  "recipe": {
    "version": 1,
    "kind": "seedRange",
    "label": "seed",
    "parameters": {
      "key": "seed",
      "start": 10,
      "count": 4,
      "step": 2
    }
  },
  "note": null
}

Recipe

{
  "version": 1,
  "kind": "seedRange",
  "label": "seed",
  "parameters": {
    "key": "seed",
    "start": 10,
    "count": 4,
    "step": 2
  }
}

Common uses

Create seed candidates for random nodes in one batch.
Switch through them one at a time with VariantSelect.
Batch-produce results by cycling seeds with BatchRender.

Notes

Generated values follow start + step * index.
Setting key to a variable name such as seed is usually easiest to understand.

Output Ports