GPT Image 2

Use the model ID gpt-image-2 with the generation or editing endpoint. Select the Direct resource group before sending requests.

GPT Image 2 returns Base64-encoded image data in data[].b64_json. Do not send the DALL·E-only response_format parameter. Select the tab for your workflow.

Generate images
Transparent output
Edit images
Mask edits

Generate images

POST /v1/images/generations

cURL

curl --request POST \
  --url https://www.anyfast.ai/v1/images/generations \
  --header 'Authorization: Bearer YOUR_API_KEY' \
  --header 'Content-Type: application/json' \
  --data '{
    "model": "gpt-image-2",
    "prompt": "A product photograph of a cobalt glass bottle on pale stone, soft morning light",
    "size": "1024x1024",
    "quality": "low",
    "output_format": "png",
    "moderation": "auto",
    "n": 1
  }'

Request headers

string

required

Bearer authentication in the form Bearer YOUR_API_KEY.

string

default:"application/json"

required

Data exchange format.

Request body

string

required

Model ID. Must be gpt-image-2.

string

required

Description of the image to generate. Length: 1–32,000 characters.

integer

default:1

Number of images to generate. Supported range: 1–10.

string

default:"auto"

auto or custom WIDTHxHEIGHT dimensions. Width and height must be divisible by 16, the aspect ratio must be from 1:3 through 3:1, neither edge may exceed 3,840 pixels, and total pixels must be from 655,360 through 8,294,400. Resolutions above 2560x1440 are experimental.

string

Rendering quality: low, medium, or high.

string

default:"auto"

Background mode: auto, opaque, or transparent. Transparent output is in preview and requires PNG or WebP.

string

default:"png"

Output format: png, jpeg, or webp.

integer

default:100

JPEG or WebP compression level from 0 through 100. Ignored for PNG output.

string

default:"auto"

Moderation strictness: auto or low.

Generation response

200

{
  "created": 1789052391,
  "data": [
    {
      "b64_json": "iVBORw0KGgoAAA..."
    }
  ],
  "model": "gpt-image-2",
  "usage": {
    "input_tokens": 24,
    "input_tokens_details": {
      "image_tokens": 0,
      "text_tokens": 24
    },
    "output_tokens": 196,
    "output_tokens_details": {
      "image_tokens": 196,
      "text_tokens": 0
    },
    "total_tokens": 220
  }
}

integer

Creation time as a Unix timestamp in seconds.

object[]

Generated image results.

string

Base64-encoded image data.

string

Model that produced the response.

object

Token usage for the request.

integer

Total text and image input tokens.

object

Input token breakdown by text and image.

integer

Image input tokens.

integer

Text input tokens.

integer

Total output tokens.

object

Output token breakdown by text and image.

integer

Image output tokens.

integer

Text output tokens.

integer

Total input and output tokens.

Transparent output

POST /v1/images/generations

cURL

curl --request POST \
  --url https://www.anyfast.ai/v1/images/generations \
  --header 'Authorization: Bearer YOUR_API_KEY' \
  --header 'Content-Type: application/json' \
  --data '{
    "model": "gpt-image-2",
    "prompt": "A simple green circular app icon, isolated on a fully transparent canvas, no shadow",
    "size": "1024x1024",
    "quality": "low",
    "background": "transparent",
    "output_format": "png",
    "n": 1
  }'

Request headers

string

required

Bearer authentication in the form Bearer YOUR_API_KEY.

string

default:"application/json"

required

Data exchange format.

Request body

string

required

Model ID. Must be gpt-image-2.

string

required

Description of the image to generate. Length: 1–32,000 characters.

integer

default:1

Number of images to generate. Supported range: 1–10.

string

default:"auto"

string

Rendering quality: low, medium, or high.

string

default:"auto"

Set to transparent for this capability. Transparent-background output is a preview feature for GPT Image 2.

string

default:"png"

Use png or webp for transparent output. The default is png. JPEG does not support transparency.

integer

default:100

JPEG or WebP compression level from 0 through 100. Ignored for PNG output.

string

default:"auto"

Moderation strictness: auto or low.

Transparent output response

200

{
  "created": 1789052391,
  "data": [
    {
      "b64_json": "iVBORw0KGgoAAA..."
    }
  ],
  "model": "gpt-image-2",
  "usage": {
    "input_tokens": 24,
    "input_tokens_details": {
      "image_tokens": 0,
      "text_tokens": 24
    },
    "output_tokens": 196,
    "output_tokens_details": {
      "image_tokens": 196,
      "text_tokens": 0
    },
    "total_tokens": 220
  }
}

integer

Creation time as a Unix timestamp in seconds.

object[]

Generated image results.

string

Base64-encoded image data.

string

Model that produced the response.

object

Token usage for the request.

integer

Total text and image input tokens.

object

Input token breakdown by text and image.

integer

Image input tokens.

integer

Text input tokens.

integer

Total output tokens.

object

Output token breakdown by text and image.

integer

Image output tokens.

integer

Text output tokens.

integer

Total input and output tokens.

Edit images

POST /v1/images/edits

cURL

curl --request POST \
  --url https://www.anyfast.ai/v1/images/edits \
  --header 'Authorization: Bearer YOUR_API_KEY' \
  --form 'model=gpt-image-2' \
  --form 'image[]=@/path/to/product.png' \
  --form 'image[]=@/path/to/reference.png' \
  --form 'prompt=Keep the product unchanged and apply the lighting and color palette from the reference image' \
  --form 'size=1536x1024' \
  --form 'output_format=png' \
  --form 'n=1'

Request headers

string

required

Bearer authentication in the form Bearer YOUR_API_KEY.

string

default:"multipart/form-data"

required

Set automatically by multipart-aware clients such as cURL.

Request body

Provide either image or image[] as the image input.

string

required

Model ID. Must be gpt-image-2.

file

One PNG or JPEG source image, smaller than 50 MB. Use image[] instead when uploading multiple images.

file[]

As many as 16 PNG or JPEG source images, each smaller than 50 MB. Do not send both image and image[].

string

required

Instructions describing the edit. Length: 1–32,000 characters.

integer

default:1

Number of edited images to return. Supported range: 1–10.

string

default:"auto"

string

default:"png"

Output format: png, jpeg, or webp.

integer

default:100

JPEG or WebP compression level from 0 through 100. Ignored for PNG output.

Edit response

200

{
  "created": 1789052450,
  "data": [
    {
      "b64_json": "iVBORw0KGgoAAA..."
    }
  ]
}

integer

Creation time as a Unix timestamp in seconds.

object[]

Edited image results.

string

Base64-encoded image data.

GPT Image 2 automatically processes edit inputs at high fidelity. Omit input_fidelity; the model does not allow changing it.

Mask edits

POST /v1/images/edits

cURL

curl --request POST \
  --url https://www.anyfast.ai/v1/images/edits \
  --header 'Authorization: Bearer YOUR_API_KEY' \
  --form 'model=gpt-image-2' \
  --form 'image=@/path/to/source.png' \
  --form 'mask=@/path/to/mask.png' \
  --form 'prompt=Keep the product unchanged and replace only the masked background with a warm studio gradient' \
  --form 'size=1536x1024' \
  --form 'output_format=png' \
  --form 'n=1'

Request headers

string

required

Bearer authentication in the form Bearer YOUR_API_KEY.

string

default:"multipart/form-data"

required

Set automatically by multipart-aware clients such as cURL.

Request body

Provide either image or image[] as the image input. The mask guides the edit area and does not guarantee a pixel-exact boundary.

string

required

Model ID. Must be gpt-image-2.

file

One PNG or JPEG source image, smaller than 50 MB. Use image[] instead when uploading multiple images.

file[]

As many as 16 PNG or JPEG source images, each smaller than 50 MB. Do not send both image and image[].

file

required

Required for mask editing; optional on the general edit endpoint. The mask must contain an alpha channel, match the first source image’s format and dimensions, and be smaller than 50 MB. Transparent regions guide the edit. With multiple source images, the mask applies to the first image.

string

required

Instructions describing the edit. Length: 1–32,000 characters.

integer

default:1

Number of edited images to return. Supported range: 1–10.

string

default:"auto"

string

default:"png"

Output format: png, jpeg, or webp.

integer

default:100

JPEG or WebP compression level from 0 through 100. Ignored for PNG output.

Mask edit response

200

{
  "created": 1789052450,
  "data": [
    {
      "b64_json": "iVBORw0KGgoAAA..."
    }
  ]
}

integer

Creation time as a Unix timestamp in seconds.

object[]

Edited image results.

string

Base64-encoded image data.

GPT Image 2 automatically processes edit inputs at high fidelity. Omit input_fidelity; the model does not allow changing it.

Validation rules

background: "transparent" requires output_format: "png" or output_format: "webp".
Custom dimensions must satisfy the divisibility, aspect-ratio, edge-length, and total-pixel limits described above.
output_compression applies only to JPEG and WebP output.
GPT Image 2 returns Base64 image data; URL output through response_format is not supported.

See the OpenAI GPT Image 2 model page, create image reference, and create image edit reference for the upstream contract.

Overview

Native Compatible Endpoints

Model APIs

Generate images

Request headers

Request body

Generation response

Transparent output

Request headers

Request body

Transparent output response

Edit images

Request headers

Request body

Edit response

Mask edits

Request headers

Request body

Mask edit response

Validation rules

​Generate images

​Request headers

​Request body

​Generation response

​Transparent output

​Request headers

​Request body

​Transparent output response

​Edit images

​Request headers

​Request body

​Edit response

​Mask edits

​Request headers

​Request body

​Mask edit response

​Validation rules

Generate images

Request headers

Request body

Generation response

Transparent output

Request headers

Request body

Transparent output response

Edit images

Request headers

Request body

Edit response

Mask edits

Request headers

Request body

Mask edit response

Validation rules