GPT-Image-2 Official Channel Image Generation

Generate images with the official OpenAI gpt-image-2 model — text-to-image, image-to-image, and inpainting (mask) in one API, with 1K / 2K / 4K resolution tiers

POST /v1/images/generations

Text-to-Image

curl --request POST \
  --url https://toapis.com/v1/images/generations \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
    "model": "gpt-image-1.5-official",
    "prompt": "Ancient castle under the starry sky, cinematic lighting, hyper-realistic style",
    "size": "16:9",
    "resolution": "2k",
    "quality": "high",
    "n": 1
  }'

200

{
  "id": "tsk_img_01KPTXXXXXXXXXXXXXXX",
  "object": "generation.task",
  "model": "gpt-image-1.5-official",
  "status": "queued",
  "progress": 0,
  "created_at": 1703884800,
  "metadata": {}
}

• Official OpenAI gpt-image-2 model, called via the gpt-image-1.5-official model name
• Async processing — returns a task ID for subsequent polling
• Supports text-to-image, multi-reference image-to-image, and mask-based inpainting
• 13 aspect ratios with 1K / 2K / 4K resolution tiers
• Up to 4 images per request, up to 16 reference images

Note: gpt-image-1.5-official does not support transparent backgrounds. Passing background: "transparent" will be silently downgraded to auto.

Authorizations

Authorization · string · required

All endpoints require Bearer Token authentication

Get your API Key from the API Key Management Page

Add to request header:

Authorization: Bearer YOUR_API_KEY

Body

model · string · default: "gpt-image-1.5-official" · required

Image generation model name

Use gpt-image-1.5-official

prompt · string · required

Text prompt describing the image to generate

Supports English and Chinese; detailed descriptions of scene, style, and composition are recommended

size · string · default: "1:1"

Aspect ratio

Supported values, or pass auto to let the server choose:

1:1 · 3:2 · 2:3 · 4:3 · 3:4 · 5:4 · 4:5 · 16:9 · 9:16 · 2:1 · 1:2 · 21:9 · 9:21

4K tier only supports 6 ratios: 16:9 / 9:16 / 2:1 / 1:2 / 21:9 / 9:21. Use 2K for all other ratios.

resolution · string · default: "1k"

Resolution tier

• 1k — 1024-base, fast and cost-effective (default)
• 2k — 2048-base, suitable for posters and high-definition needs
• 4k — 3840-base, only supports the 6 ratios above; high quality can take >120s

Size Reference Table

size	1K	2K	4K
1:1	1024x1024	2048x2048	Not supported
3:2	1536x1024	2048x1360	Not supported
2:3	1024x1536	1360x2048	Not supported
4:3	1024x768	2048x1536	Not supported
3:4	768x1024	1536x2048	Not supported
5:4	1280x1024	2560x2048	Not supported
4:5	1024x1280	2048x2560	Not supported
16:9	1536x864	2048x1152	3840x2160
9:16	864x1536	1152x2048	2160x3840
2:1	2048x1024	2688x1344	3840x1920
1:2	1024x2048	1344x2688	1920x3840
21:9	2016x864	2688x1152	3840x1648
9:21	864x2016	1152x2688	1648x3840

quality · string · default: "high"

Image quality

• low — fast and economical, suitable for drafts/previews
• medium — balanced speed and quality
• high — maximum detail, default value (4K + high can take 120s+)

output_format · string · default: "png"

Output format

• png — default
• jpeg — smaller file size (supports compression)

Azure OpenAI does not support webp format.

output_compression · integer · default: 100

JPEG compression level, range 0–100 (0 = no compression, 100 = maximum compression, default 100)

Only applies when output_format is "jpeg"

n · integer · default: 1

Number of images to generate

Range: 1 ~ 10

image_urls · string[]

Reference image URL array for image-to-image generation

• Maximum 16 images, must be publicly accessible stable URLs
• Use the Upload Image API to upload local files and get URLs

mask_url · string

Mask image URL for inpainting

Must be used together with image_urls. The mask must match the dimensions of the first reference image and include an Alpha channel (transparent areas indicate regions to be repainted)

Response

id · string

Unique task identifier for status queries

object · string

Object type, always generation.task

model · string

Model name used

status · string

Task status

• queued — waiting to be processed
• in_progress — processing
• completed — successfully completed
• failed — failed

progress · integer

Task progress percentage (0-100)

created_at · integer

Task creation timestamp (Unix timestamp)

Text-to-Image

curl --request POST \
  --url https://toapis.com/v1/images/generations \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
    "model": "gpt-image-1.5-official",
    "prompt": "Ancient castle under the starry sky, cinematic lighting, hyper-realistic style",
    "size": "16:9",
    "resolution": "2k",
    "quality": "high",
    "n": 1
  }'

200

{
  "id": "tsk_img_01KPTXXXXXXXXXXXXXXX",
  "object": "generation.task",
  "model": "gpt-image-1.5-official",
  "status": "queued",
  "progress": 0,
  "created_at": 1703884800,
  "metadata": {}
}