API Documentation

Use the AIGPTImage API to create image and video generation tasks from your backend.

The API is asynchronous by default. A successful create request returns a taskId; your server then polls the task status endpoint until the result is ready.

What you can do

Create image generation tasks.
Create video generation tasks.
Query task status and final result URLs.
Reuse the same credit balance and pricing rules as the web app.

Base URL

https://aigptimage.com

Use https://aigptimage.com as the production API base URL.

Authentication

Create an API key from the API Keys page, then click Create API Key.

Send it with every request:

Authorization: Bearer sk-your-api-key

Core endpoints

Endpoint	Method	Purpose
`/api/v1/jobs/createTask`	`POST`	Create an image or video generation task
`/api/v1/jobs/recordInfo`	`GET`	Query a task by `taskId`

Supported models

Model ID	Type	Scene	Required input	Market path	Docs
gpt-image-2-text-to-image	image	Text to image	prompt	/market/gpt/gpt-image-2-text-to-image.md	Open
gpt-image-2-image-to-image	image	Image to image	prompt, image_urls	/market/gpt/gpt-image-2-image-to-image.md	Open
seedance-2.0-text-to-video	video	Text to video	prompt	/market/seedance/seedance-2.0-text-to-video.md	Open
seedance-2.0-image-to-video	video	Image to video	prompt, image_urls	/market/seedance/seedance-2.0-image-to-video.md	Open

Credit consumption rules

Public API calls use the same credit balance and pricing engine as the web app.

Current public models

Public model	Media type	Billing rule	Current credit cost
`gpt-image-2-text-to-image`	Image	Fixed credits per output, based on `resolution`	`1K` = 3 credits, `2K` = 5 credits, `4K` = 8 credits
`gpt-image-2-image-to-image`	Image	Fixed credits per output, based on `resolution`	`1K` = 3 credits, `2K` = 5 credits, `4K` = 8 credits
`seedance-2.0-text-to-video`	Video	Credits per second, based on `resolution` and `duration`	`480p` = 20/s, `720p` = 40/s, `1080p` = 80/s
`seedance-2.0-image-to-video`	Video	Credits per second, based on `resolution` and `duration`	`480p` = 20/s, `720p` = 40/s, `1080p` = 80/s

GPT Image 2 pricing

GPT Image 2 image generation uses a fixed-per-output pricing rule:

1K: 3 credits per image
2K: 5 credits per image
4K: 8 credits per image

This applies to both:

gpt-image-2-text-to-image
gpt-image-2-image-to-image

Seedance 2.0 pricing

Seedance 2.0 video generation uses a per-second pricing rule. Total credits depend on:

resolution
duration

Resolution	Credits per second
`480p`	20
`720p`	40
`1080p`	80

This default mapping currently applies to both:

seedance-2.0-text-to-video
seedance-2.0-image-to-video

Default calculation examples

seedance-2.0-text-to-video, 720p, 5s: 40 x 5 = 200 credits
seedance-2.0-image-to-video, 1080p, 5s: 80 x 5 = 400 credits

Important notes

API pricing is shared with the web app. There is no separate API-only credit wallet.
The final credit cost is calculated when the task is created.
If a request is invalid and the task is not created, credits are not consumed.
If a task fails after credit reservation, the existing refund logic is applied.

Response format

All public API responses use the same shape:

{
  "code": 200,
  "msg": "success",
  "data": {}
}

Errors use the same format with a non-200 code and HTTP status.

Next step

Start with Getting Started, then integrate Create Task and Query Task.

API Documentation

On this page