Quick Start
Get an API key and complete your first API call in 2 minutes
Quick Start
This guide walks you through signing up, topping up credits, creating an API Key, and making your first API request with cURL. Most users complete their first call within 5 minutes.
Prerequisites
- Network access to
https://api.icreat.ai - A terminal (macOS, Linux, or Windows PowerShell)
- Optional: cURL and jq for requests and JSON parsing
1. Sign Up
Visit iCreat AI and register using one of the following methods:
Option 1: Google sign-in
- Click Sign in with Google
- Authorize with your Google account
- Your account is created automatically and ready to use
Option 2: Email registration
Go to the registration page to create an iCreat account:
- Enter your username and email
- Set a password (8–20 characters)
- Click Get Verification Code and enter the code from your email (check spam if you do not receive it)
- Submit to complete registration Already have an account? Sign in to open the console.
Tip
Keep your login credentials safe. We recommend using a strong password.
2. Top Up Credits
iCreat uses a prepaid credit model. API usage is deducted from your balance. After logging in, open Wallet to add credits:
- Per-call pricing varies by model — see the Model Plaza
- When balance is insufficient, the API returns an error — top up promptly
- View balance and usage history in the console
Transparent billing
Each API call is logged in real time. You can filter usage by API Key and model.
3. Create API Keys
Your API Key authenticates requests via Authorization: Bearer <API_KEY>.
On the API Keys page:
- Click Add
- Enter a name (e.g.
production-backend,local-dev) - Optionally set a credit cap and expiration
- Submit
Security
- Keys are shown in full only once at creation — copy and store them securely
- Never expose keys in client-side code or public repositories
- Revoke and rotate immediately if a key is leaked
- Use separate keys per environment and app for audit and rotation See API Key management for details.
4. Your First API Call
Image and video generation uses an async two-step flow:
Submit task → Query task result (poll until terminal)The example below uses GPT-Image-2 text-to-image. Replace <your-api-key> with your key and reuse the same task_id in both steps.
4.1 Submit a task
POST https://api.icreat.ai/v1/task/submit/openai/gpt-image-2
curl -X POST 'https://api.icreat.ai/v1/task/submit/openai/gpt-image-2' \
-H 'Authorization: Bearer <your-api-key>' \
-H 'Content-Type: application/json' \
-d '{
"prompt": "a Shiba Inu wearing sunglasses, cyberpunk style",
"size": "1024x1024",
"quality": "high"
}'Response example:
{
"task_id": "task-0192xxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"
}Save task_id for the next step.
4.2 Query task result (poll)
POST https://api.icreat.ai/v1/task/result
Call this endpoint with the same task_id until status is SUCCEEDED or FAILED. The body includes status and result: result is [] while processing; on success, result holds resources and costUSD is present.
curl -X POST 'https://api.icreat.ai/v1/task/result' \
-H 'Authorization: Bearer <your-api-key>' \
-H 'Content-Type: application/json' \
-d '{"task_id": "task-0192xxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx"}'While processing:
{
"status": "SUBMITTED",
"result": []
}On success:
{
"status": "SUCCEEDED",
"costUSD": 0.0123,
"result": [
{
"type": "Image",
"url": "https://cdn.example.com/generated/xxx.png",
"download_url": "https://cdn.example.com/generated/xxx.png?download=1"
}
]
}Use result[0].url to view the image and download_url to download. If status is FAILED, review your request and submit again.
Polling tip
If the task is not terminal yet, wait 2–5 seconds and call
/v1/task/resultagain. Avoid excessive polling.
Next steps
- API overview — LLM, image, and video model docs
- Integration guides — Claude Code, Codex, and other tools
