---
title: GPT Image 1.5 | Runware Docs
url: https://runware.ai/docs/models/openai-gpt-image-1-5
description: GPT Image 1.5 flagship image model with faster generation and enhanced editing
---
# GPT Image 1.5

![GPT Image 1.5](https://assets.runware.ai/covers/openai-gpt-image-1-5.jpg)

GPT Image 1.5 is OpenAI’s newest flagship image model powering the latest ChatGPT Images. It delivers significantly faster image generation with stronger instruction following, more precise edits that preserve original details, more believable transformations, and improved rendering of dense or small text. It is suited for practical creative workflows, detailed design tasks, and production use cases.

- **Status**: deprecated
- **ID**: `openai:4@1`
- Try in Playground
- **Creator**: OpenAI
- **Release Date**: December 16, 2025
- **Capabilities**: Text to Image, Image to Image, Edit, Checkpoint, Transparency

## Pricing

~$0.136 typical run

**Rates** (per 1M tokens)

- Text input: `$5.00`
- Image input: `$8.00`
- Image output: `$32.00`

**4 example generations**

- [reference-to-image](https://runware.ai/docs/models/openai-gpt-image-1-5/examples#example-ferry-crew-archive-restoration): `$0.0575` (~28s)
- [text-to-image](https://runware.ai/docs/models/openai-gpt-image-1-5/examples#example-quilted-lineworker-archive-portrait): `$0.135` (~37s)
- [text-to-image](https://runware.ai/docs/models/openai-gpt-image-1-5/examples#example-marine-emergency-signal-kit-packshot): `$0.136` (~33s)
- [inpaint](https://runware.ai/docs/models/openai-gpt-image-1-5/examples#example-wildlife-hospital-cut-paper-artwork): `$0.204` (~48s)

billed per token, so a request costs what the model produced · times vary, measured end to end · [How pricing works](https://runware.ai/docs/models-api/pricing)

## Compatibility & Validation

When `inputs.maskImage` is provided, `inputs.referenceImages` is required.

---

When `settings.background` is `transparent`, `outputFormat` is limited to `PNG`, `WEBP`.

---

`width` and `height` must be used together.

---

The following dimension combinations are supported:

| Configuration | Dimensions |
| --- | --- |
| `1K (1:1)` | `1024x1024` |
| `1K (3:2)` | `1536x1024` |
| `1K (2:3)` | `1024x1536` |

## Request Parameters

**API Options**

Platform-level options for task execution and delivery.

### [taskType](https://runware.ai/docs/models/openai-gpt-image-1-5#request-tasktype)

- **Type**: `string`
- **Required**: true
- **Value**: `imageInference`

Identifier for the type of task being performed.

### [taskUUID](https://runware.ai/docs/models/openai-gpt-image-1-5#request-taskuuid)

- **Type**: `string`
- **Required**: true
- **Format**: `UUID v4`

UUID v4 identifier for tracking tasks and matching async responses. Must be unique per task.

### [outputType](https://runware.ai/docs/models/openai-gpt-image-1-5#request-outputtype)

- **Type**: `string`
- **Default**: `URL`

Image output type.

**Allowed values**: `URL` `base64Data` `dataURI`

### [outputFormat](https://runware.ai/docs/models/openai-gpt-image-1-5#request-outputformat)

- **Type**: `string`
- **Required**: conditional — Required when settings.background is transparent
- **Default**: `JPG`

Specifies the file format of the generated output. The available values depend on the task type and the specific model's capabilities.

- \`JPG\`: Best for photorealistic images with smaller file sizes (no transparency).
- \`PNG\`: Lossless compression, supports high quality and transparency (alpha channel).
- \`WEBP\`: Modern format providing superior compression and transparency support.

> [!NOTE]
> \*\*Transparency\*\*: If you are using features like background removal or LayerDiffuse that require transparency, you must select a format that supports an alpha channel (e.g., \`PNG\`, \`WEBP\`, \`TIFF\`). \`JPG\` does not support transparency.

**Allowed values**: `JPG` `PNG` `WEBP`

### [outputQuality](https://runware.ai/docs/models/openai-gpt-image-1-5#request-outputquality)

- **Type**: `integer`
- **Min**: `20`
- **Max**: `99`
- **Default**: `95`

Compression quality of the output. Higher values preserve quality but increase file size.

### [webhookURL](https://runware.ai/docs/models/openai-gpt-image-1-5#request-webhookurl)

- **Type**: `string`
- **Format**: `uri`

Webhook URL that receives JSON responses via HTTP POST when generation tasks complete. For batch requests with multiple results, each completed item triggers a separate webhook call as it becomes available.

**Learn more** (1 resource):

- [Webhooks](https://runware.ai/docs/models-api/webhooks) (platform)

### [deliveryMethod](https://runware.ai/docs/models/openai-gpt-image-1-5#request-deliverymethod)

- **Type**: `string`
- **Default**: `sync`

Determines how the API delivers task results.

**Allowed values**:

- `sync` Returns results directly in the API response. Over HTTP, a task still running after 120 seconds continues as an async task and comes back with status 'processing'. Poll for its results using getResponse.
- `async` Returns an immediate acknowledgment with the task UUID. Poll for results using getResponse.

**Learn more** (1 resource):

- [Task Polling](https://runware.ai/docs/models-api/task-polling) (platform)

### [uploadEndpoint](https://runware.ai/docs/models/openai-gpt-image-1-5#request-uploadendpoint)

- **Type**: `string`
- **Format**: `uri`

Specifies a URL where the generated content will be automatically uploaded using the HTTP PUT method. The raw binary data of the media file is sent directly as the request body. For secure uploads to cloud storage, use presigned URLs that include temporary authentication credentials.

**Common use cases:**

- **Cloud storage**: Upload directly to S3 buckets, Google Cloud Storage, or Azure Blob Storage using presigned URLs.
- **CDN integration**: Upload to content delivery networks for immediate distribution.

```text
// S3 presigned URL for secure upload
https://your-bucket.s3.amazonaws.com/generated/content.mp4?X-Amz-Signature=abc123&X-Amz-Expires=3600

// Google Cloud Storage presigned URL
https://storage.googleapis.com/your-bucket/content.jpg?X-Goog-Signature=xyz789

// Custom storage endpoint
https://storage.example.com/uploads/generated-image.jpg
```

The content data will be sent as the request body to the specified URL when generation is complete.

### [safety](https://runware.ai/docs/models/openai-gpt-image-1-5#request-safety)

- **Path**: `safety.checkContent`
- **Type**: `object (1 property)`

Content safety checking configuration for image generation.

#### [checkContent](https://runware.ai/docs/models/openai-gpt-image-1-5#request-safety-checkcontent)

- **Path**: `safety.checkContent`
- **Type**: `boolean`

Enable or disable content safety checking. Increases total generation time and adds a small charge on top of the generation cost.

### [ttl](https://runware.ai/docs/models/openai-gpt-image-1-5#request-ttl)

- **Type**: `integer`
- **Min**: `60`

Time-to-live (TTL) in seconds for generated content. Only applies when `outputType` is `URL`.

### [includeCost](https://runware.ai/docs/models/openai-gpt-image-1-5#request-includecost)

- **Type**: `boolean`

Include task cost in the response.

### [numberResults](https://runware.ai/docs/models/openai-gpt-image-1-5#request-numberresults)

- **Type**: `integer`
- **Min**: `1`
- **Max**: `20`
- **Default**: `1`

Number of results to generate. Each result uses a different seed, producing variations of the same parameters.

**Inputs**

Input resources for the task (images, audio, etc). These must be nested inside the \`inputs\` object.

### [maskImage](https://runware.ai/docs/models/openai-gpt-image-1-5#request-inputs-maskimage)

- **Path**: `inputs.maskImage`
- **Type**: `string`

Image used to specify which areas of the input image should be edited (UUID, URL, Data URI, or Base64). Must have the same width and height as the first reference image. White marks the area to edit, black is preserved. The model treats the mask as guidance rather than a hard boundary, so edits can extend past its shape. With several reference images the mask applies to the first.

### [referenceImages](https://runware.ai/docs/models/openai-gpt-image-1-5#request-inputs-referenceimages)

- **Path**: `inputs.referenceImages`
- **Type**: `array of strings`
- **Min items**: `1`
- **Max items**: `16`

List of reference images (UUID, URL, Data URI, or Base64).

**Core Parameters**

Primary parameters that define the task output.

### [model](https://runware.ai/docs/models/openai-gpt-image-1-5#request-model)

- **Type**: `string`
- **Required**: true
- **Value**: `openai:4@1`

Send `auto` instead of an AIR identifier to have the model router choose the model for this task. The rest of the request is unchanged, every value you send is kept, and only the parameters you leave out are filled in. See [Model Router](https://runware.ai/docs/models-api/model-router).

**Learn more** (3 resources):

- [Text To Image: Model Selection](https://runware.ai/docs/learn/text-to-image#model-selection) (learn)
- [Image Inpainting: Model Specialized Inpainting Models](https://runware.ai/docs/learn/image-inpainting#model-specialized-inpainting-models) (learn)
- [Image Outpainting: Other Critical Parameters](https://runware.ai/docs/learn/image-outpainting#other-critical-parameters) (learn)

### [positivePrompt](https://runware.ai/docs/models/openai-gpt-image-1-5#request-positiveprompt)

- **Type**: `string`
- **Required**: true
- **Min**: `2`
- **Max**: `32000`

Text prompt describing elements to include in the generated output.

**Learn more** (1 resource):

- [Prompts](https://runware.ai/docs/learn/prompts) (learn)

### [width](https://runware.ai/docs/models/openai-gpt-image-1-5#request-width)

- **Type**: `integer`
- **Paired with**: height

Width of the generated media in pixels.

**Learn more** (2 resources):

- [Dimensions](https://runware.ai/docs/learn/dimensions) (learn)
- [Image Outpainting: Dimensions Critical For Outpainting](https://runware.ai/docs/learn/image-outpainting#dimensions-critical-for-outpainting) (learn)

### [height](https://runware.ai/docs/models/openai-gpt-image-1-5#request-height)

- **Type**: `integer`
- **Paired with**: width

Height of the generated media in pixels.

**Learn more** (2 resources):

- [Dimensions](https://runware.ai/docs/learn/dimensions) (learn)
- [Image Outpainting: Dimensions Critical For Outpainting](https://runware.ai/docs/learn/image-outpainting#dimensions-critical-for-outpainting) (learn)

**Settings**

Technical parameters to fine-tune the inference process. These must be nested inside the \`settings\` object.

### [background](https://runware.ai/docs/models/openai-gpt-image-1-5#request-settings-background)

- **Path**: `settings.background`
- **Type**: `string`
- **Default**: `auto`

Background handling in generated images.

**Allowed values**:

- `auto` Automatically determines the most appropriate background based on content.
- `opaque` Solid background, no transparency.
- `transparent` Transparent background areas where applicable.

**Features**

Standalone addons and post-processing features.

### [watermark](https://runware.ai/docs/models/openai-gpt-image-1-5#request-advancedfeatures-watermark)

- **Path**: `advancedFeatures.watermark`
- **Type**: `object (6 properties)`

Configuration object for adding watermarks to generated videos. Watermarks can be applied using either text or image content with customizable positioning and appearance. You must provide either `text` or `image` content for the watermark, but not both.

**Text watermark**:

```json
"advancedFeatures": {
  "watermark": {
    "text": "© 2025 Company",
    "displayPosition": "bottom-right",
    "opacity": 0.6,
    "fontColor": "#ffffff",
    "bgColor": "#000000"
  }
}
```

**Image watermark**:

```json
"advancedFeatures": {
  "watermark": {
    "image": "c64351d5-4c59-42f7-95e1-eace013eddab",
    "displayPosition": "top-left",
    "opacity": 0.6
  }
}
```

**Tiled watermark**:

```json
"advancedFeatures": {
  "watermark": {
    "text": "PREVIEW",
    "tiled": true,
    "opacity": 0.4,
    "fontColor": "#cccccc"
  }
}
```

#### [text](https://runware.ai/docs/models/openai-gpt-image-1-5#request-advancedfeatures-watermark-text)

- **Path**: `advancedFeatures.watermark.text`
- **Type**: `string`
- **Min**: `2`
- **Max**: `32`

Watermark text.

#### [image](https://runware.ai/docs/models/openai-gpt-image-1-5#request-advancedfeatures-watermark-image)

- **Path**: `advancedFeatures.watermark.image`
- **Type**: `string`

Watermark image (UUID, URL, Data URI, or Base64).

#### [displayPosition](https://runware.ai/docs/models/openai-gpt-image-1-5#request-advancedfeatures-watermark-displayposition)

- **Path**: `advancedFeatures.watermark.displayPosition`
- **Type**: `string`

Watermark position.

**Allowed values**: `top-left` `top-center` `top-right` `center-left` `center-center` `center-right` `bottom-left` `bottom-center` `bottom-right` `tiled`

#### [opacity](https://runware.ai/docs/models/openai-gpt-image-1-5#request-advancedfeatures-watermark-opacity)

- **Path**: `advancedFeatures.watermark.opacity`
- **Type**: `float`
- **Min**: `0.1`
- **Max**: `1`
- **Step**: `0.01`

Watermark opacity.

#### [fontColor](https://runware.ai/docs/models/openai-gpt-image-1-5#request-advancedfeatures-watermark-fontcolor)

- **Path**: `advancedFeatures.watermark.fontColor`
- **Type**: `string`

Text color in hex format.

#### [bgColor](https://runware.ai/docs/models/openai-gpt-image-1-5#request-advancedfeatures-watermark-bgcolor)

- **Path**: `advancedFeatures.watermark.bgColor`
- **Type**: `string`

Background color in hex format.

**Provider Settings**

Parameters specific to this model provider. These must be nested inside the \`providerSettings.openai\` object.

### [moderation](https://runware.ai/docs/models/openai-gpt-image-1-5#request-providersettings-openai-moderation)

- **Path**: `providerSettings.openai.moderation`
- **Type**: `string`
- **Default**: `auto`

Content moderation level.

**Allowed values**:

- `auto` Standard filtering that limits creating certain categories of potentially age-inappropriate content.
- `low` Less restrictive filtering.

### [quality](https://runware.ai/docs/models/openai-gpt-image-1-5#request-providersettings-openai-quality)

- **Path**: `providerSettings.openai.quality`
- **Type**: `string`
- **Default**: `auto`

Image quality level.

**Allowed values**: `auto` `high` `medium` `low`

## Response Parameters

### [taskType](https://runware.ai/docs/models/openai-gpt-image-1-5#response-tasktype)

- **Type**: `string`
- **Required**: true
- **Value**: `imageInference`

Identifier for the type of task this response belongs to.

### [taskUUID](https://runware.ai/docs/models/openai-gpt-image-1-5#response-taskuuid)

- **Type**: `string`
- **Required**: true
- **Format**: `UUID v4`

UUID v4 identifier echoed from the original request, used to match async responses to their tasks.

### [imageUUID](https://runware.ai/docs/models/openai-gpt-image-1-5#response-imageuuid)

- **Type**: `string`
- **Required**: true
- **Format**: `UUID v4`

UUID of the output image.

### [imageURL](https://runware.ai/docs/models/openai-gpt-image-1-5#response-imageurl)

- **Type**: `string`
- **Format**: `uri`

URL of the output image.

### [imageBase64Data](https://runware.ai/docs/models/openai-gpt-image-1-5#response-imagebase64data)

- **Type**: `string`

Base64-encoded image data.

### [imageDataURI](https://runware.ai/docs/models/openai-gpt-image-1-5#response-imagedatauri)

- **Type**: `string`
- **Format**: `uri`

Data URI of the output image.

### [NSFWContent](https://runware.ai/docs/models/openai-gpt-image-1-5#response-nsfwcontent)

- **Type**: `boolean`

Flag indicating if NSFW content was detected.

### [cost](https://runware.ai/docs/models/openai-gpt-image-1-5#response-cost)

- **Type**: `float`

Task cost in USD. Present when `includeCost` is set to `true` in the request.

## Examples

### text-to-image (text-to-image)

![Marine Emergency Signal Kit Packshot](https://assets.runware.ai/examples/openai-gpt-image-1-5/b2e21d36-f945-4091-88b5-cc77c32dbdee.jpg)

```json
{
  "taskType": "imageInference",
  "taskUUID": "d9ba2631-25f9-4107-9d9d-81484e1075a5",
  "model": "openai:4@1",
  "positivePrompt": "Create a premium ecommerce product packshot of a complete marine emergency signaling kit. Center an open, rugged safety-orange waterproof carrying case in a three-quarter front view, with charcoal-gray latches, a black rubber perimeter seal, a sturdy molded handle, and precisely fitted dark navy closed-cell foam inside. Arrange the contents clearly and logically: two cylindrical red hand flares, one compact orange smoke signal canister, a navy LED distress beacon with a translucent lens, an orange signal whistle on a short black cord, and a neatly coiled reflective rescue line. Keep several items secured in their foam cutouts and place one hand flare, the beacon, and the whistle upright in front of the case so shoppers can inspect their forms. Every object must be complete, correctly proportioned, physically plausible, and visually distinct, with clean manufacturing details and no duplicated accessories.\n\nUse a restrained industrial palette of safety orange, signal red, dark navy, charcoal, and brushed metal. Photograph the kit against a seamless warm light-gray studio background with a soft floor gradient. High-key commercial lighting: a large diffused key from upper left, gentle fill from the front, crisp controlled rim light along the case edges, realistic contact shadows, subtle reflections on plastic and metal, and excellent texture definition without harsh glare. Square composition with generous even margins, product occupying roughly 78 percent of the frame, camera at product height with a slightly elevated view so both the interior organization and front-facing items are visible. Sharp focus throughout, realistic product photography, accurate materials, polished catalogue retouching, neutral color balance, no dramatic environment and no decorative props.\n\nRender only these prominent generic labels, exactly as written and clearly legible: on the case lid, “OFFSHORE SIGNAL KIT” with the smaller line “FOR MARINE EMERGENCY USE”; on each red cylinder, “RED HAND FLARE”; on the orange canister, “ORANGE SMOKE SIGNAL”; on the beacon, “LED DISTRESS BEACON” and a small “SOS” button label. Use clean condensed sans-serif safety typography printed directly on the products. Do not invent a company name, brand logo, certification mark, website, price, barcode, promotional badge, or additional copy. The final image must be a single isolated ecommerce packshot, not packaging artwork, not a lifestyle scene, and not a website or app mockup.",
  "width": 1024,
  "height": 1024,
  "providerSettings": {
    "openai": {
      "quality": "high"
    }
  }
}
```

---

### text-to-image (text-to-image)

![Quilted Lineworker Archive Portrait](https://assets.runware.ai/examples/openai-gpt-image-1-5/a4a9e3ac-bed2-4f3d-9f45-847d7f826c75.jpg)

```json
{
  "taskType": "imageInference",
  "taskUUID": "b4dcb0b5-f12d-4f39-aab7-b1fe33e41c07",
  "model": "openai:4@1",
  "positivePrompt": "Create a finished square commemorative artwork that looks like a 1970s documentary photograph transformed into a handmade American story quilt. Depict five rural electrical lineworkers—three women and two men—standing beside a weathered bucket truck after repairing storm damage. Preserve the convincing composition and candid specificity of an archival crew photo: full figures arranged in a loose row, one worker resting a gloved hand on a coiled cable, another holding a hard hat at her hip, utility poles and repaired power lines receding across flat farmland behind them. Their expressions are tired, proud, and understated rather than heroic poses.\n\nRender every element as layered fabric appliqué, dense hand embroidery, visible running stitches, quilted seams, frayed denim edges, corduroy, canvas, wool felt, and tiny thread highlights. Faces must remain recognizable and anatomically coherent despite the textile treatment. Construct the bucket truck, insulators, tool belts, work boots, gloves, cable, and pole hardware with accurate, readable shapes. Use a restrained heritage palette of faded indigo, workwear tan, rust orange, mustard yellow, brick red, warm cream, and charcoal, with subtle variations from aged cloth. The cloudy post-storm sky should be assembled from broad overlapping fabric panels, while fine embroidered lines describe rain, electrical wires, and windblown grass.\n\nFront-facing, balanced square composition with the crew occupying the central two-thirds and enough environmental context to establish rural utility work. Soft overcast lighting translated into fabric value changes; tactile shadows cast by raised seams and appliqué layers. Authentic handmade imperfections, but polished enough for a museum exhibition graphic. No frame, wall, room, mockup, logos, captions, borders, watermarks, or added typography; the textile artwork itself fills the entire image.",
  "width": 1024,
  "height": 1024,
  "providerSettings": {
    "openai": {
      "quality": "high"
    }
  }
}
```

---

### reference-to-image (reference-to-image)

![Ferry Crew Archive Restoration](https://assets.runware.ai/examples/openai-gpt-image-1-5/ea2080b1-1108-4d5a-99b9-72c614f7d238.jpg)

```json
{
  "taskType": "imageInference",
  "taskUUID": "40431a09-c7d2-4cca-8d34-a0297f604ef0",
  "model": "openai:4@1",
  "positivePrompt": "Create a faithful, museum-quality restoration and colorization of reference image 1. Reference image 1 is the authoritative source for the entire composition, people, facial identities, poses, clothing shapes, tools, ferry structure, scaffolding, camera angle, crop, depth of field, and documentary character. Repair its diagonal tear, creases, scratches, dust, water stains, fading, and missing emulsion; reconstruct only the small details logically interrupted by that physical damage. Preserve every visible person and object in the same location and do not beautify, modernize, replace, remove, or add anyone.\n\nUse reference image 2 only as the historical color and material guide: apply its restrained late-1970s palette of deep navy coveralls, mustard-yellow hard hats, pale seafoam-green and cream ferry paint, safety-orange railings, blue-gray steel, red welding leads, wet charcoal concrete, and muted evergreen background tones to the corresponding elements in reference image 1. Match the overcast Pacific Northwest light and believable surface qualities shown in reference image 2, including worn paint, oxidized steel, cotton workwear, and damp pavement. Do not copy the second image's composition or introduce objects from it.\n\nThe finished image should look like the original 1978 photograph had survived in excellent condition and had been captured on authentic color negative film: natural skin tones across the crew, subtle tonal separation, controlled saturation, fine period-appropriate grain, gentle highlight roll-off, and modest edge softness. Retain small signs of age such as very fine grain and slight optical softness while removing distracting physical damage. Avoid contemporary digital sharpness, HDR contrast, glossy AI smoothing, invented lettering, altered faces, duplicated people, anachronistic safety gear, or cinematic relighting. Deliver one clean full-frame landscape archival photograph with no frame, mat, caption, watermark, comparison layout, or before-and-after split.",
  "width": 1536,
  "height": 1024,
  "providerSettings": {
    "openai": {
      "quality": "medium"
    }
  },
  "inputs": {
    "referenceImages": [
      "https://assets.runware.ai/assets/inputs/af5b02ef-9913-47d4-8b97-ba8557d86db4.jpg",
      "https://assets.runware.ai/assets/inputs/e1a04f30-d57c-4067-a8bf-e154b7835e2c.jpg"
    ]
  }
}
```

---

### inpaint (inpaint)

![Wildlife Hospital Cut-Paper Artwork](https://assets.runware.ai/examples/openai-gpt-image-1-5/73c24415-d15d-41a0-80ba-44e4dc319f16.jpg)

```json
{
  "taskType": "imageInference",
  "taskUUID": "90b60ca8-7d91-4526-95c8-3cb20e6ab4b4",
  "model": "openai:4@1",
  "positivePrompt": "Edit the supplied photograph into finished annual report cover artwork for a wildlife rehabilitation hospital. Replace the entire barn owl with a convincing handcrafted, dimensional cut-paper interpretation of the same owl. Preserve its exact species, upright pose, proportions, folded-wing silhouette, head angle, placement, scale, feet position, and alert forward gaze. Construct the owl from many precisely cut and subtly raised paper layers: a cream heart-shaped facial disc, charcoal and warm-brown feather markings, rust-red wing accents, muted mustard highlights, and restrained deep-teal details. Use elegant simplified feather shapes inspired by mid-century natural-history illustration, with visible paper fibers, crisp cut edges, tiny registration variations, and delicate shadows between overlapping layers. Keep the eyes expressive and anatomically aligned, rendered as layered black and amber paper rather than photographic eyes. The result should feel like an expertly photographed physical paper sculpture occupying the owl's original position, not a flat digital filter. Preserve the wooden perch, seamless warm-beige background, original lighting direction, negative space, camera angle, depth of field, and all surrounding scene details unchanged. Maintain believable contact where the paper feet wrap around the perch and add only the subtle shadows needed to integrate the transformed owl naturally into the existing photograph. Sophisticated editorial craft aesthetic, calm and compassionate mood, clean portrait composition, no typography, no logo, no border, no additional objects.",
  "width": 1024,
  "height": 1536,
  "providerSettings": {
    "openai": {
      "quality": "high"
    }
  },
  "inputs": {
    "referenceImages": [
      "https://assets.runware.ai/assets/inputs/5ce06ee5-fac1-42eb-91fa-ab03b185aceb.jpg"
    ],
    "maskImage": "https://assets.runware.ai/assets/inputs/1403b241-4046-4f70-9e98-f2bb0a7a2864.png"
  }
}
```