GPT Image 2.5 Sunburst API: Image Generation & Editing

Generate and edit images with the GPT Image 2.5 Sunburst API through APIXO

Overview

GPT Image 2.5 Sunburst is an asynchronous image generation API for text-to-image and reference-guided image editing.

Capability	Value
Model ID	gpt-image-2.5-sunburst
Modes	text-to-image, image-to-image
Prompt length	1-32,000 characters
Reference images	1-16 URLs for image-to-image
Mask editing	Optional mask_url for image-to-image
Aspect ratios	auto, 1:1, 1:2, 2:1, 2:3, 3:2, 4:3, 3:4, 4:5, 5:4, 16:9, 9:16, 21:9
Output control	aspect_ratio or size (supported ratio or custom WIDTHxHEIGHT)
Background	auto, transparent, opaque
Output format	png, jpeg, webp
Resolution tiers	1k, 2k, 4k
Quality tiers	low, medium, high, xhigh, max

Endpoint and authentication

Base URL:

https://api.apixo.ai/api/v1

Method	Endpoint	Purpose
POST	/generateTask/gpt-image-2.5-sunburst	Submit a generation task
GET	/statusTask/gpt-image-2.5-sunburst?taskId={taskId}	Poll task status and retrieve results

Authorization: Bearer YOUR_API_KEY
Content-Type: application/json

Quickstart

curl -X POST "https://api.apixo.ai/api/v1/generateTask/gpt-image-2.5-sunburst" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "request_type": "async",
    "input": {
      "mode": "text-to-image",
      "prompt": "A cinematic aerial view of a futuristic coastal city at sunrise, warm golden light, detailed architecture, realistic clouds and ocean, no text or watermark",
      "aspect_ratio": "16:9",
      "resolution": "4k",
      "quality": "max"
    }
  }'

The accepted response includes data.taskId. Save it and poll until state becomes success or failed.

Request body

Text-to-image

{
  "request_type": "async",
  "input": {
    "mode": "text-to-image",
    "prompt": "A cinematic travel poster of mountains, a lake, and a cabin at sunrise, rich natural colors, no text",
    "aspect_ratio": "3:2",
    "resolution": "2k",
    "quality": "high"
  }
}

Image-to-image

{
  "request_type": "async",
  "input": {
    "mode": "image-to-image",
    "prompt": "Turn the supplied portrait into a refined editorial fashion photograph. Preserve the person's identity and pose, natural skin texture, beige studio backdrop, soft directional light.",
    "image_urls": [
      "https://your-domain.com/images/portrait-source.jpg"
    ],
    "mask_url": "https://your-domain.com/images/portrait-mask.png",
    "aspect_ratio": "4:5",
    "background": "opaque",
    "output_format": "png",
    "resolution": "2k",
    "quality": "max"
  }
}

Parameters

request_type · string · default: "async" · required

Result delivery mode. Use async for polling with statusTask, or callback for webhook delivery.

callback_url · string

Required when request_type is callback. It must be a public HTTPS URL that can receive the final task payload. See Webhooks.

input · object · required

GPT Image 2.5 Sunburst input parameters.

Show properties

mode · string · required

Generation mode. Supported values: text-to-image, image-to-image.

prompt · string · required

Text prompt describing the requested image or edit. Supports 1-32,000 characters.

image_urls · string[]

Publicly accessible reference image URLs. Required for image-to-image; supports 1-16 URLs.

aspect_ratio · string · default: "auto"

Output aspect ratio. Supported values: auto, 1:1, 1:2, 2:1, 2:3, 3:2, 4:3, 3:4, 4:5, 5:4, 16:9, 9:16, 21:9. 1:2 and 2:1 require resolution to be 2k or 4k. Ignored when size is supplied.

size · string

Optional output size. Use a supported aspect ratio or custom WIDTHxHEIGHT, such as 1536x1024. It takes priority over aspect_ratio. Custom WIDTHxHEIGHT, 1:2, and 2:1 require resolution to be 2k or 4k.

resolution · string · default: "1k"

Output resolution tier. Supported values: 1k, 2k, 4k.

quality · string · default: "medium"

Quality and billing tier. Supported values: low, medium, high, xhigh, max. Each tier has its own listed price.

background · string · default: "auto"

Output background. Supported values: auto, transparent, opaque. When using transparent, set output_format to png or webp.

output_format · string · default: "png"

Output image format. Supported values: png, jpeg, webp. jpeg cannot be used with background=transparent.

mask_url · string

Optional public HTTP(S) mask image URL. Available only for image-to-image and requires image_urls.

Poll for results

curl -X GET "https://api.apixo.ai/api/v1/statusTask/gpt-image-2.5-sunburst?taskId=task_12345678" \
  -H "Authorization: Bearer YOUR_API_KEY"

When successful, data.resultJson is a JSON string containing resultUrls:

{
  "code": 200,
  "message": "success",
  "data": {
    "taskId": "task_12345678",
    "state": "success",
    "resultJson": "{\"resultUrls\":[\"https://file.apixo.ai/xxx.png\"]}",
    "createTime": 1767965610929,
    "completeTime": 1767965652317,
    "costTime": 41388
  }
}

For failed tasks, state is failed and the response includes failCode and failMsg.

Billing

Billing is per successfully generated image. Prices below are in USD and retain four decimal places. Text-to-image and image-to-image use the same output-tier price; reference images do not add an extra charge.

Quality	1k	2k	4k
low	$0.0100	$0.0200	$0.0300
medium	$0.0240	$0.0400	$0.0700
high	$0.0900	$0.1500	$0.2700
xhigh	$0.1600	$0.2700	$0.4800
max	$0.3600	$0.6000	$1.0000

For the live price catalog, see Pricing.

Latency and polling

Generation time depends on prompt complexity, reference images, selected resolution, and current queue load. Start polling after 30-60 seconds, then poll every 10 seconds. For production workloads, prefer callback mode.

Related links

• Image Models
• Generation API Overview
• Generate Task
• Status Task
• Webhooks
• Pricing