> ## Documentation Index
> Fetch the complete documentation index at: https://docs.sprello.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Image Input

> Hold one or more images to use as references or sources.

An Image Input node holds the images you bring to a workflow—sketches, product photos, material swatches, or brand references—to guide what a generator produces.

## Add images

* **Upload** a file from your computer.
* **Paste** an image directly onto the node.
* **Drag** an image onto the canvas to create an Image Input.
* **Pick from your [asset library](/organize/asset-library)** to reuse something you already have.

An Image Input can hold a **single image or many**.

<Tip>
  Open an image from here to [annotate or inpaint](/concepts/annotate-and-inpaint) it—mark it up, or regenerate a single region—before feeding it to a generator.
</Tip>

## Each or All

When an Image Input holds several images, the [Each / All toggle](/concepts/variations) decides how they're used:

* **Each** runs the generator once per image—five product photos become five separate generations.
* **All** sends the images together as one set of references—three brand references shaping a single result.

See [generating variations](/concepts/variations) for the full picture.

## Connect it

Where an Image Input connects depends on the generator:

* **Generate Image** and **Generate Text** accept it as **reference images**.
* **Generate Video** uses it as a **source frame**.

The number of references a generator accepts depends on its model—see [connections and data types](/concepts/connections).