Gemini Image Generation

Overview

AIone supports Gemini image generation models, accessible through the standard OpenAI-compatible /v1/chat/completions endpoint.

These models are internally routed through the Gemini native image generation pipeline, with responses converted back to the OpenAI-compatible format. Your request format stays the same, but you can use additional image-specific parameters.

Available Image Models

The following Gemini image models are currently available:

gemini-3-pro-image-preview

gemini-3-pro-image-preview-2k

gemini-3-pro-image-preview-4k

gemini-2.5-flash-image

gemini-3.1-flash-image-preview

If you are unsure about a model name, refer to GET https://api.nexara.net/v1/models and the Portal model list page.

Minimal Example

Notes:

aspect_ratio and image_size are optional -- images will be generated without them

The response uses the standard OpenAI chat.completion structure

Example with Aspect Ratio and Size Parameters

Supported Request Parameters

General Parameters

model: Image model ID

messages: Conversation message array, following the standard OpenAI Chat Completions format

max_tokens / max_completion_tokens: Mapped to Gemini's native maxOutputTokens

temperature

top_p

top_k

Image-Specific Parameters

aspect_ratio: Mapped to Gemini's native imageConfig.aspectRatio

image_size: Mapped to Gemini's native imageConfig.imageSize

Both parameters are optional -- images will generate without them.

Reference Image Input

To include a reference image, use the OpenAI-style multimodal messages.content format:

{
  "model": "gemini-3-pro-image-preview",
  "messages": [
    {
      "role": "user",
      "content": [
        {"type": "text", "text": "Using this image as a composition reference, draw a cat sitting by a window"},
        {
          "type": "image_url",
          "image_url": {
            "url": "data:image/png;base64,iVBORw0KGgoAAA..."
          }
        }
      ]
    }
  ],
  "max_tokens": 4096
}

Notes:

The gateway currently only supports data: URI format for image_url.url on the Gemini native pipeline

If you pass a public image URL directly, the gateway will not convert it to a Gemini native image input

Response Format

Image model responses may return choices[0].message.content as an array rather than a plain string:

{
  "id": "chatcmpl-xxx",
  "object": "chat.completion",
  "model": "gemini-3-pro-image-preview",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": [
          {"type": "text", "text": "Here is the generated image:"},
          {
            "type": "image_url",
            "image_url": {
              "url": "data:image/png;base64,AAAA..."
            }
          }
        ]
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 123,
    "completion_tokens": 456,
    "total_tokens": 579
  }
}

This means:

Some responses include both descriptive text and an image

Image content is returned via image_url.url, typically in data:image/...;base64,... format

The usage field is still returned, allowing you to cross-reference consumption in both the response and the console

Limitations & Notes

Only /v1/chat/completions is supported: Gemini image models do not support /v1/messages

Streaming is not supported: stream must be false or omitted

Response content may be an array: Do not assume choices[0].message.content is always a plain text string

Model permissions: Ensure your API Key is authorized to access the image model

Model variants: If you specifically want the 2K or 4K variant, use gemini-3-pro-image-preview-2k or gemini-3-pro-image-preview-4k directly

14 - Gemini Image Generation

Gemini Image Generation#

Overview#

Available Image Models#

Minimal Example#

Example with Aspect Ratio and Size Parameters#

Supported Request Parameters#

General Parameters#

Image-Specific Parameters#

Reference Image Input#

Response Format#

Limitations & Notes#