Create image

💡 Troubleshooting

Running into an error while calling the API? See the API error code reference for a complete list of error types and troubleshooting steps.

POST https://zenmux.ai/api/v1/images/generations

The Create image endpoint is compatible with OpenAI's Create image endpoint and creates images from prompts.

The parameters below cover what different models may support. Support varies by model. ZenMux passes through the OpenAI Images API protocol; when calling ZenMux, set the Base URL to https://zenmux.ai/api/v1 and authenticate with your ZenMux API Key.

Currently Supported Models

ZenMux continuously updates the OpenAI image generation models it supports. Visit the ZenMux model catalog to check the currently available models.

Request headers

Authorization `string`

Bearer Token authentication, formatted as Bearer $ZENMUX_API_KEY.

Content-Type `string`

The request content type. Defaults to application/json.

Request body

prompt `string`

Text description of the image to generate. GPT image models support a maximum length of 32,000 characters.

background `string`

Sets the background transparency of the generated image. This parameter is only supported by GPT image models. Must be one of transparent, opaque, or auto; defaults to auto. With auto, the model automatically determines the most suitable background for the image.

If set to transparent, the output format must support transparency, so output_format should be set to png (default) or webp.

Possible values:

transparent
opaque
auto

model `string`

The model to use for image generation. Visit the ZenMux model catalog to check the currently available models.

moderation `string`

Controls the content moderation level for GPT image model generation. Must be one of low or auto (default).

Possible values:

low
auto

n `number`

The number of images to generate. Must be between 1 and 10.

output_compression `number`

The compression level for generated images (0-100%). This parameter is only supported by GPT image models with webp or jpeg output formats, and defaults to 100.

output_format `string`

The format used to return generated images. This parameter is only supported by GPT image models. Must be one of png, jpeg, or webp.

Possible values:

png
jpeg
webp

partial_images `number`

The number of partial images to generate. This parameter is used for streaming responses that return partial images. The value must be between 0 and 3. When set to 0, the response returns a single image in one streaming event.

If the complete image is generated faster, the final image may be sent before all requested partial images have been produced.

quality `string`

The quality of the generated image. ZenMux currently supports GPT image model quality parameters for OpenAI image generation models.

auto (default) automatically selects the best quality for the given model.
high, medium, and low are supported by GPT image models.

Possible values:

low
medium
high
auto

response_format `string`

The return format for the generated image. This parameter does not apply to the GPT image models currently supported by ZenMux; GPT image models always return base64-encoded images.

Possible values:

url
b64_json

size `string`

The size of the generated image. For gpt-image-2 and gpt-image-2-2026-04-21, arbitrary WIDTHxHEIGHT strings are supported, such as 1536x864. Both width and height must be divisible by 16, and the requested aspect ratio must be between 1:3 and 3:1. Resolutions above 2560x1440 are experimental, with a maximum supported resolution of 3840x2160. Requested sizes must also satisfy the model's current pixel and side-length limits.

gpt-image-1.5 supports the standard sizes 1024x1024, 1536x1024, and 1024x1536; models that support automatic sizing may use auto.

Common possible values:

auto
1024x1024
1536x1024
1024x1536

stream `boolean`

Generate the image in streaming mode. Defaults to false. This parameter is only supported by GPT image models.

style `string`

The style of the generated image. This parameter does not apply to the OpenAI image generation models currently supported by ZenMux.

Possible values:

vivid
natural

user `string`

A unique identifier representing your end user, which can help with monitoring and abuse detection.

Response

Returns the response object from the image generation endpoint.

created `number`

The Unix timestamp (in seconds) when the image was created.

background `string`

The background parameter used during image generation. May be transparent or opaque.

data `array`

The list of generated images.

b64_json `string`

The base64-encoded JSON for the generated image. GPT image models currently supported by ZenMux return this field by default.

revised_prompt `string`

Returned by some models. Represents the rewritten prompt actually used to generate the image.

url `string`

May contain a generated image URL for models that return URLs. GPT image models currently supported by ZenMux usually return b64_json.

output_format `string`

The output format of the generated image. May be png, webp, or jpeg.

quality `string`

The generated image quality. May be low, medium, or high.

size `string`

The generated image size. May be 1024x1024, 1024x1536, or 1536x1024.

usage `object`

Token usage information for GPT image model image generation.

input_tokens number: The number of image and text tokens in the input prompt.
input_tokens_details object: Input token details.
- image_tokens number: The number of image tokens in the input prompt.
- text_tokens number: The number of text tokens in the input prompt.
output_tokens number: The number of output tokens generated by the model.
total_tokens number: The total number of tokens used for this image generation.
output_tokens_details object : Output token details.
- image_tokens number: The number of generated image output tokens.
- text_tokens number: The number of generated text output tokens.

cURLTypeScriptPython

cURL

curl https://zenmux.ai/api/v1/images/generations \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $ZENMUX_API_KEY" \
  -d '{
    "model": "gpt-image-2",
    "prompt": "A cute baby sea otter",
    "n": 1,
    "size": "1024x1024"
  }'

TypeScript

import OpenAI from "openai";
import { writeFile } from "fs/promises";

const client = new OpenAI({
  baseURL: "https://zenmux.ai/api/v1", 
  apiKey: process.env.ZENMUX_API_KEY, 
});

const img = await client.images.generate({
  model: "gpt-image-2",
  prompt: "A cute baby sea otter",
  n: 1,
  size: "1024x1024",
});

const imageBuffer = Buffer.from(img.data[0].b64_json, "base64");
await writeFile("output.png", imageBuffer);

Python

import base64
from openai import OpenAI

client = OpenAI(
    base_url="https://zenmux.ai/api/v1",  
    api_key="<ZENMUX_API_KEY>",  
)

img = client.images.generate(
    model="gpt-image-2",
    prompt="A cute baby sea otter",
    n=1,
    size="1024x1024"
)

image_bytes = base64.b64decode(img.data[0].b64_json)
with open("output.png", "wb") as f:
    f.write(image_bytes)

json

{
  "created": 1713833628,
  "data": [
    {
      "b64_json": "..."
    }
  ],
  "usage": {
    "total_tokens": 100,
    "input_tokens": 50,
    "output_tokens": 50,
    "input_tokens_details": {
      "text_tokens": 10,
      "image_tokens": 40
    }
  }
}

Create image ​

Request headers ​

Authorization string ​

Content-Type string ​

Request body ​

prompt string ​

background string ​

model string ​

moderation string ​

n number ​

output_compression number ​

output_format string ​

partial_images number ​

quality string ​

response_format string ​

size string ​

stream boolean ​

style string ​

user string ​

Response ​

created number ​

background string ​

data array ​

b64_json string ​

revised_prompt string ​

url string ​

output_format string ​

quality string ​

size string ​

usage object ​

Create image

Request headers

Authorization `string`

Content-Type `string`

Request body

prompt `string`

background `string`

model `string`

moderation `string`

n `number`

output_compression `number`

output_format `string`

partial_images `number`

quality `string`

response_format `string`

size `string`

stream `boolean`

style `string`

user `string`

Response

created `number`

background `string`

data `array`

b64_json `string`

revised_prompt `string`

url `string`

output_format `string`

quality `string`

size `string`

usage `object`