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

# Image nodes

> Generate and edit images in Creative Studio — text-to-image, image-to-image, upscale, inpaint, background removal, reframe, and SVG vectorization with Auto Router and SOTA models.

Image nodes are the foundation of Creative Studio's visual pipeline. Each node runs one image operation: pick a **mode**, choose a **model** (or leave it on **Auto**), set parameters, and generate. Chain nodes to build multi-step editing flows.

<CardGroup cols={2}>
  <Card title="Generate" icon="wand-sparkles">
    Text to image, image to image, and SVG creation from prompts or references.
  </Card>

  <Card title="Edit" icon="pen-nib">
    Upscale, reframe, inpaint, erase, and background tools on existing assets.
  </Card>
</CardGroup>

## Auto Router & the model picker

Every image node has a model selector. Leave it on **Auto** to let the router choose the best model for your prompt and mode, or open the picker to choose one yourself — each option shows its speed and credit cost.

<Steps>
  <Step title="Auto routes for you" icon="sparkles">
    Auto reads your prompt, mode, and constraints, then routes to the model that best balances quality, speed, and cost for that request.

    <Frame>
      <img className="rounded-xl" style={{ maxWidth: "420px", width: "100%" }} src="https://mintcdn.com/elarislabs-87/7tt1Idyze_8Pr4_1/images/studio/auto-router.png?fit=max&auto=format&n=7tt1Idyze_8Pr4_1&q=85&s=2958ce4eaae6add33ba114c20bdccb11" alt="Image node set to Auto with a detailed generation prompt" width="990" height="946" data-path="images/studio/auto-router.png" />
    </Frame>
  </Step>

  <Step title="Or pick a model yourself" icon="layers">
    Open the picker to search compatible models, jump into featured options like **Flux** or **GPT Image**, or browse providers — Google, Black Forest Labs, ByteDance, OpenAI, Kling, Runway, and more.

    <Frame>
      <img className="rounded-xl" style={{ maxWidth: "420px", width: "100%" }} src="https://mintcdn.com/elarislabs-87/7tt1Idyze_8Pr4_1/images/studio/model-picker.jpg?fit=max&auto=format&n=7tt1Idyze_8Pr4_1&q=85&s=1d34f88e347b0d124aed42f1272a1cf9" alt="Model picker showing featured models and providers with credit costs" width="824" height="1024" data-path="images/studio/model-picker.jpg" />
    </Frame>
  </Step>
</Steps>

Creative Studio ships **state-of-the-art models across domains and price points** — photorealism, fine-grained edits, fast drafts, typography, and SVG. Full capability matrix: [Image models](/docs/models/image).

<Tip>
  Resolution and quality change credit cost. Higher resolutions (2K / 4K) and premium models cost more per run — check the estimate in the node panel before generating.
</Tip>

## What generation modes are available?

Create new images from text or from existing references.

### Text to image

Generate a new image from a text prompt. Set aspect ratio, resolution, and optional style references. Works with Auto Router and all major models.

**Best for:** hero shots, concept art, product visuals, ad backgrounds.

<Frame>
  <video className="w-full aspect-video rounded-xl" autoPlay loop muted playsInline src="https://mintcdn.com/elarislabs-87/qcKqm6Hw9hLc3Cdj/images/studio/text-to-image.mp4?fit=max&auto=format&n=qcKqm6Hw9hLc3Cdj&q=85&s=c3b2d3f98ad5a62466bb07eb4249aaed" data-path="images/studio/text-to-image.mp4" />
</Frame>

### Image to image

Condition generation on one or more reference images. Control how closely the output follows the reference via a strength / influence slider.

**Best for:** style transfer, iterating on existing shots, brand-consistent generation.

<Frame>
  <video className="w-full aspect-video rounded-xl" autoPlay loop muted playsInline src="https://mintcdn.com/elarislabs-87/qcKqm6Hw9hLc3Cdj/images/studio/image-to-image.mp4?fit=max&auto=format&n=qcKqm6Hw9hLc3Cdj&q=85&s=ccb186adee11725bceecc30e10e16473" data-path="images/studio/image-to-image.mp4" />
</Frame>

## What editing modes are available?

Transform an existing image without starting from scratch.

<Tabs>
  <Tab title="Upscale">
    Increase resolution and add fine detail. Useful as a final step before export or handoff.

    **Input:** any image node output or uploaded asset.

    <Frame>
      <video className="w-full aspect-video rounded-xl" autoPlay loop muted playsInline src="https://mintcdn.com/elarislabs-87/D0vwEQ0svpFEhVwG/images/studio/upscale.mp4?fit=max&auto=format&n=D0vwEQ0svpFEhVwG&q=85&s=e73186a2e06cabebf263d87db80d7c3c" data-path="images/studio/upscale.mp4" />
    </Frame>
  </Tab>

  <Tab title="Reframe">
    Outpaint an image to a new aspect ratio — expand the canvas on any edge. Ideal for adapting a portrait crop to landscape without regenerating.

    **Input:** any image.

    <Frame>
      <video className="w-full aspect-video rounded-xl" autoPlay loop muted playsInline src="https://mintcdn.com/elarislabs-87/D0vwEQ0svpFEhVwG/images/studio/reframe.mp4?fit=max&auto=format&n=D0vwEQ0svpFEhVwG&q=85&s=4507c89825e5d7bef6f53910b38beb43" data-path="images/studio/reframe.mp4" />
    </Frame>

    Once reframed, hand the image to **[Brand Pack](/docs/studio/brand-pack)** to resize across every placement — Stories, feed, landscape DOOH — in one pass.

    <Frame>
      <video className="w-full aspect-video rounded-xl" autoPlay loop muted playsInline src="https://mintcdn.com/elarislabs-87/D0vwEQ0svpFEhVwG/images/studio/brand-pack.mp4?fit=max&auto=format&n=D0vwEQ0svpFEhVwG&q=85&s=d442ffd66ecbe528c871ef49f19a0bc9" data-path="images/studio/brand-pack.mp4" />
    </Frame>
  </Tab>

  <Tab title="Mask / inpaint">
    Draw a mask over a region and describe what should replace it. The rest of the image stays untouched.

    **Input:** source image + mask drawing.

    <Frame>
      <video className="w-full aspect-video rounded-xl" autoPlay loop muted playsInline src="https://mintcdn.com/elarislabs-87/D0vwEQ0svpFEhVwG/images/studio/inpaint.mp4?fit=max&auto=format&n=D0vwEQ0svpFEhVwG&q=85&s=4b1cd5a2db9f4077de0936528f6496f5" data-path="images/studio/inpaint.mp4" />
    </Frame>
  </Tab>

  <Tab title="Erase">
    Remove an unwanted object or region and fill it with the surrounding background naturally.

    **Input:** source image + selection.

    <Frame>
      <video className="w-full aspect-video rounded-xl" autoPlay loop muted playsInline src="https://mintcdn.com/elarislabs-87/D0vwEQ0svpFEhVwG/images/studio/erase.mp4?fit=max&auto=format&n=D0vwEQ0svpFEhVwG&q=85&s=0a027c1a062452df24bd2d2ecdb4633b" data-path="images/studio/erase.mp4" />
    </Frame>
  </Tab>

  <Tab title="Replace background">
    Automatically segment the subject and swap the background with a new scene described by prompt.

    **Input:** source image + prompt for the new background.

    <Frame>
      <video className="w-full aspect-video rounded-xl" autoPlay loop muted playsInline src="https://mintcdn.com/elarislabs-87/D0vwEQ0svpFEhVwG/images/studio/replace-background.mp4?fit=max&auto=format&n=D0vwEQ0svpFEhVwG&q=85&s=9334473ccab77bf8056042c63f782163" data-path="images/studio/replace-background.mp4" />
    </Frame>
  </Tab>

  <Tab title="Remove background">
    Cut out the subject and export with a transparent background (PNG alpha). Common before the layer editor or brand pack.

    **Input:** any image.

    <Frame>
      <video className="w-full aspect-video rounded-xl" autoPlay loop muted playsInline src="https://mintcdn.com/elarislabs-87/D0vwEQ0svpFEhVwG/images/studio/remove-background.mp4?fit=max&auto=format&n=D0vwEQ0svpFEhVwG&q=85&s=dbbe729ac9e8b2a08d7c506ca1d084ca" data-path="images/studio/remove-background.mp4" />
    </Frame>
  </Tab>
</Tabs>

## What SVG / vector modes are available?

Generate or vectorize artwork as clean SVG paths — useful for logos, icons, and print-ready assets.

<Frame>
  <video className="w-full aspect-video rounded-xl" autoPlay loop muted playsInline src="https://mintcdn.com/elarislabs-87/qcKqm6Hw9hLc3Cdj/images/studio/vectorise.mp4?fit=max&auto=format&n=qcKqm6Hw9hLc3Cdj&q=85&s=5b64f8b3e33346c9f154bd62a1cc16b5" data-path="images/studio/vectorise.mp4" />
</Frame>

<CardGroup cols={2}>
  <Card title="Text to SVG" icon="pen-tool">
    Generate scalable vector artwork directly from a text description using Quiver Arrow.
  </Card>

  <Card title="Image to SVG" icon="scan-line">
    Vectorize any raster image into clean SVG paths for logos, icons, and print.
  </Card>
</CardGroup>

## Which models are available?

Open the model picker on any image node for featured options and full provider lists. Common families include:

| Model / Family           | Strengths                                      |
| ------------------------ | ---------------------------------------------- |
| **Flux**                 | High-quality photorealism and prompt adherence |
| **Nano Banana (Gemini)** | Fast generation, strong at text-in-image       |
| **Seedream**             | Stylized and artistic outputs                  |
| **GPT Image**            | Instruction-following, multi-subject scenes    |
| **Kling**                | Cinematic quality, strong composition          |
| **Wan**                  | Versatile general-purpose generation           |
| **Runway**               | Creative / editorial style                     |
| **Ideogram utilities**   | Typography and design                          |
| **Phota**                | Portrait and product photography style         |
| **Grok Imagine**         | Expressive and imaginative outputs             |
| **Hunyuan**              | Asian aesthetic, high detail                   |
| **Quiver Arrow**         | SVG vector generation                          |

Full matrix: [Image models](/docs/models/image).

## What's next?

<CardGroup cols={3}>
  <Card title="Video nodes" icon="video" href="/docs/studio/video-nodes">
    Animate your images or generate from scratch
  </Card>

  <Card title="Layer editor" icon="layers" href="/docs/studio/layer-editor">
    Composite multiple images into one layout
  </Card>

  <Card title="Brand pack" icon="package" href="/docs/studio/brand-pack">
    Resize your image for every platform
  </Card>
</CardGroup>