Ship AI Image Generation in Your App — GPT Image 2 API

Free GPT Image 2 API Access

Every new account receives free credits automatically — no credit card required. Your first API calls cost nothing. Test the gpt image 2 api, explore the endpoints, and verify output quality before committing to a paid plan.

Credit-Based Billing

API usage consumes the same credits as the web app. Credits never expire within their validity window. Failed tasks before acceptance automatically refund your credits — you only pay for successful images.

POST /api/v1/images/generations — Create Image

GET /api/v1/images/tasks/{task_id} — Query Task

GPT-4 Image API Input — image_urls Parameter

For image-to-image workflows (similar to gpt-4o api image input or gpt 4.1 image input api), pass one or more reference image URLs via the image_urls array. The GPT Image 2 model will use these as visual references while applying your text prompt. This enables product photography retouching, style transfer, background replacement, and precise object editing without reinterpreting the entire scene.

JavaScript / Node.js — Full Polling Example

This exact pattern runs in production apps today. Copy it, swap the prompt and model, it works.

const apiKey = process.env.GPT_IMAGE_2_API_KEY;

// Step 1: Create image generation task
const createRes = await fetch(
  'https://gpt-image-2.art/api/v1/images/generations',
  {
    method: 'POST',
    headers: {
      Authorization: `Bearer ${apiKey}`,
      'Content-Type': 'application/json',
    },
    body: JSON.stringify({
      model: 'gpt-image-2',
      prompt: 'A cinematic product photo of a matte black coffee bag on marble',
      size: '1536x1024',
      quality: 'high',
      n: 1,
    }),
  }
);

const task = await createRes.json();
// { id: "8f6d...", status: "pending", credits: 10 }

// Step 2: Poll until succeeded
let result = task;
while (['pending', 'processing'].includes(result.status)) {
  await new Promise((r) => setTimeout(r, 3000));
  const res = await fetch(
    `https://gpt-image-2.art/api/v1/images/tasks/${task.id}`,
    { headers: { Authorization: `Bearer ${apiKey}` } }
  );
  result = await res.json();
}

// Step 3: Use the image URL
console.log(result.data[0].url); // → "https://..."

Python — Text-to-Image Example

import os, time, requests

api_key = os.environ["GPT_IMAGE_2_API_KEY"]
base_url = "https://gpt-image-2.art/api/v1"
headers = {"Authorization": f"Bearer {api_key}"}

# Create task
resp = requests.post(
    f"{base_url}/images/generations",
    headers=headers,
    json={
        "model": "gpt-image-2",
        "prompt": "Premium skincare bottle on wet stone, studio lighting",
        "size": "1024x1024",
        "quality": "high",
        "n": 1,
    },
)
task = resp.json()

# Poll for result
while task["status"] in ("pending", "processing"):
    time.sleep(3)
    task = requests.get(
        f"{base_url}/images/tasks/{task['id']}",
        headers=headers,
    ).json()

print(task["data"][0]["url"])

n8n Automation — gpt image api n8n Integration

Use the HTTP Request node in n8n to integrate the gpt image api n8n workflows. Set the method to POST, URL to https://gpt-image-2.art/api/v1/images/generations, add the Authorization header with your Bearer token, and pass your JSON body. Use a second HTTP Request node polling the task endpoint, combined with a Wait node (3s interval) inside a loop until status equals succeeded.

Complex text + diagram infographic generated in a single GPT Image 2 API call

GPT Image 2 API vs GPT-4 Image API

The gpt-4 image api refers to GPT-4 Vision's ability to accept image inputs for understanding and analysis — it does not generate images. GPT Image 2 is OpenAI's standalone image generation model. If your use case requires gpt 4o image input api or gpt 5 image input api for image understanding alongside generation, you can combine both: use GPT-4o for analysis and the GPT Image 2 API for generation within the same pipeline.

GPT Image 2 API vs Azure OpenAI Image API

The azure gpt image api routes through Microsoft's Azure infrastructure, requiring an Azure subscription, resource provisioning, and per-region availability constraints. gpt-image-2.art provides direct API access without Azure overhead — simpler authentication, faster setup, and lower per-image costs for most usage tiers.