Interface Description

To meet the strong expectations of our users, we have specially launched two reverse-engineered interfaces: gpt-4o-image and gpt-4o-image-vip. The VIP model offers superior performance and stability.
Please note that reverse-engineered interfaces cannot guarantee full stability and may encounter 🕐 Queuing... status or generation failures. They are recommended only for trial experiences and functional testing.
OpenAI’s official interface gpt-image-1 has been released. Considering the cost of usage, the current reverse-engineered interfaces will be temporarily retained, with compatible calling methods provided for everyone.

Notes

  • Special Note: Regardless of the cause, once the API call is initiated, charges will apply even if the generation fails.
  • The name of living artists (such as “Ghibli” and “Makoto Shinkai”) will trigger input_moderation error, causing generation failure. You can avoid this by using non-sensitive words like “Ghibli” and “Bright Modern Japanese Anime Style”.
  • In general, “style” is safer than “artist”, like “Pixar” is also supported.
  • A more secure approach is to use the name of deceased artists or corresponding styles, such as “Van Gogh” and “Mona Lisa”.

Model and Pricing

ModelPrice ($/IMG)
gpt-4o-image-vip$0.009
gpt-4o-image$0.005

Due to the rising maintenance costs of reverse-engineered interfaces, the price per call has been increased. (Updated on: 2025-06-04).

Python Example

Using the familiar v1/completions endpoint, we provide support, which is fully compatible. For specific usage, please refer to:

from openai import OpenAI

client = OpenAI(
    api_key="AIHUBMIX_API_KEY", # Your Key Here
    base_url="https://aihubmix.com/v1"
)

# Create a chat completion request
stream = client.chat.completions.create(
    model="gpt-4o-image-vip", #list: gpt-4o-image-vip, gpt-4o-image
    messages=[
        {
            "role": "user",
            "content": [
                {
                    "type": "text",
                    "text": """Redesign [Girl wearing a pearl earring] as an anime-style illustrated cover. Preserve original essence and composition. Render in refined Kyoto Anime-inspired style—soft brushwork, cinematic lighting, progressive depth. 
Format in AV cover layout: 
Japanese-English mixed typography, serif title with significant visual hierarchy - sense of design is the key.
Extract words tha conveys profound meanings as title, then creative slogan, the original name can be a smaller information ONLY if necessary.
Subtly embed watermark "jerlin" in clothing or background. Maintain emotional tone and posture, soften only when necessary. 2:3 portrait"""
                }
            ]
        }
    ],
    stream=True
)

# Process the streaming response
for chunk in stream:
    try:
        if chunk.choices and len(chunk.choices) > 0 and chunk.choices[0].delta and chunk.choices[0].delta.content:
            print(chunk.choices[0].delta.content, end="", flush=True)
    except Exception as e:
        print(f"\nError occurred while processing the response: {str(e)}")

Output Example

{
  "prompt": "Redesign 'Girl with a Pearl Earring' as an anime-style illustrated cover. Preserve the original essence and composition: the iconic side-glance pose, serene expression, and intimate framing. Render in refined Kyoto Anime-inspired style with soft brushwork, cinematic lighting, and progressive depth. Create an AV cover layout with Japanese-English mixed typography, using a prominent serif title with strong visual hierarchy. Extract profound words for the title and include a creative slogan. The original name 'Girl with a Pearl Earring' should appear as smaller text if needed. Subtly embed the watermark 'jerlin' in clothing or background. For modesty, replace any potential nudity with flowing fabric, hair, or veils while maintaining the emotional tone and posture. Soften details only when necessary.",
  "ratio": "2:3"
}

>🕐 Queuing.

>⚡ Generating.....

>🏃‍ Progress 11....19....27...34....42....50....57...64....[100](https://videos.openai.com/vg-assets/assets%2Ftask_01jsccq77nfcsab7dn8bx7z230%2Fsrc_1.png?st=2025-04-21T13%3A08%3A48Z&se=2025-04-27T14%3A08%3A48Z&sks=b&skt=2025-04-21T13%3A08%3A48Z&ske=2025-04-27T14%3A08%3A48Z&sktid=a48cca56-e6da-484e-a814-9c849652bcb3&skoid=aa5ddad1-c91a-4f0a-9aca-e20682cc8969&skv=2019-02-02&sv=2018-11-09&sr=b&sp=r&spr=https%2Chttp&sig=xm6K6M7au3F3Vd9qcYxR3mYWrqAHsgbagbUV5GQo39w%3D&az=oaivgprodscus)

> ✅ Generation complete


![gen_01jsccq83pfg6ssfqf3pzfaytj](https://filesystem.site/cdn/20250421/XBCrAs6qvMARG7BkFX0KI3MBs1qoto.png)

[Click to download](https://filesystem.site/cdn/download/20250421/XBCrAs6qvMARG7BkFX0KI3MBs1qoto.png)

![gen_01jsccq83nf6ftmtx8y48h383y](https://filesystem.site/cdn/20250421/tQ62gh0BbdESApNmFJMNzsdUFDflkl.png)

[Click to download](https://filesystem.site/cdn/download/20250421/tQ62gh0BbdESApNmFJMNzsdUFDflkl.png)%

Compatibility with Official OpenAI API

If you prefer using the latest image generation method released by OpenAI, you can refer to the following implementation. Just keep these 3 points in mind:

  1. Do not include the quality parameter — models like gpt-4o-image-vip and gpt-4o-image have it bundled internally.
  2. The usage field (token tracking) is not supported.
  3. The latest output returns Base64-encoded image data, and the current script already handles automatic saving.
Python
from openai import OpenAI
import os
import requests
import base64

client = OpenAI(
    api_key="AIHUBMIX_API_KEY", # Replace with your AiHubMix API key
    base_url="https://aihubmix.com/v1"
)

prompt = """
redesign poster of the movie [Black Swam], 3D cartoon, smooth render, bright tone, 2:3
"""

result = client.images.generate(
    model="gpt-4o-image-vip",
    prompt=prompt,
    n=1, # Number of images to generate
    size="1024x1536", # 1024x1024 (square), 1536x1024 (3:2 landscape), 1024x1536 (2:3 portrait), auto (default)
    #quality="high" # Do not pass quality for reverse model, it's internally bound 📍
)

print(result) # Print full API response; usage not supported for reverse API 📍

# Define output filename prefix and directory
output_dir = "." # You can specify another directory
file_prefix = "image_gen-reverse"

# Ensure output directory exists
os.makedirs(output_dir, exist_ok=True)

# Iterate through all returned image data
for i, image_data in enumerate(result.data):
    # Handle base64 string first
    if hasattr(image_data, "b64_json") and image_data.b64_json:
        image_bytes = base64.b64decode(image_data.b64_json)
        current_index = i
        while True:
            file_name = f"{file_prefix}_{current_index}.png"
            file_path = os.path.join(output_dir, file_name)
            if not os.path.exists(file_path):
                break
            current_index += 1
        with open(file_path, "wb") as f:
            f.write(image_bytes)
        print(f"Image saved to: {file_path} (base64)")
    # Then handle url field (robust fallback in case of interface change)
    elif hasattr(image_data, "url") and image_data.url:
        try:
            response = requests.get(image_data.url, stream=True)
            response.raise_for_status()
            image_bytes = response.content
            current_index = i
            while True:
                file_name = f"{file_prefix}_{current_index}.png"
                file_path = os.path.join(output_dir, file_name)
                if not os.path.exists(file_path):
                    break
                current_index += 1
            with open(file_path, "wb") as f:
                f.write(image_bytes)
            print(f"Image saved to: {file_path} (url)")
        except requests.exceptions.RequestException as e:
            print(f"Failed to download image #{i}: {e}")
        except Exception as e:
            print(f"Error occurred while processing image #{i}: {e}")
    else:
        print(f"No b64_json or url found in image #{i}, skipping save.")

Effect Demonstration

Girl-with-a-Pearl-Earring

Girl with a Pearl Earring

  1. The famous painting is reinterpreted, with text and watermark directly output.
pixar

Pixar

  1. The Pixar 3D Cartoon style, stable as ever.
ghibli

Ghibli

  1. The most popular Ghibli

Practical Tips

  • Support any language, Chinese drawing is also stable, but we don’t recommend drawing a large amount of text.
  • Specify the aspect ratio in the prompt, otherwise 4o will automatically determine it. Supports 2:3, 3:2, 1:1.
  • Give reference images if you don’t have inspiration, support single image and multi-image.
  • Upload images can be compressed for pre-processing, improve speed.
  • Support transparent background, no need to remove the background.——just add the requirement in the prompt.