Text-to-Image

hy-image-v3.5-preview generates images from text prompts via the Chat/Messages protocol, supporting custom dimensions (up to 4096x4096), seed, and watermark.

Model

Use the hy-image-v3.5-preview model with the cloudbase provider. Requires @cloudbase/js-sdk >= 3.11.0.

Parameters

Parameter	Type	Required	Description
model	string	Yes	Fixed to hy-image-v3.5-preview
messages	Array<object>	Yes	Conversation content; put the text in the content of a role=user message
size	string	No	Size in ${width}x${height} format; when set, output is forced to this size
seed	integer	No	Generation seed in [0, 2^63-1]; fix it to reproduce results
generate_max_pixels	integer	No	Area tier, effective only when size is not set
footnote	string	No	Bottom-right watermark, up to 16 characters

For the full reference (session, resize_max_pixels, use_search_tool) and response fields, see Hy-Image-3.5-preview.

Supported Sizes

Method	Description
size	Width/height in [256, 8192], area up to 4096x4096, 4K maximum
generate_max_pixels	1048576 (1K), 2359296 (1.5K, default), 4194304 (2K)
Neither	The model decides based on prompt semantics; defaults to the 1.5K tier

Usage Example

const imageModel = ai.createImageModel("cloudbase");

const res = await imageModel.generateImage({
  model: "hy-image-v3.5-preview",
  size: "4096x4096",
  seed: 42,
  footnote: "CloudBase AI",
  messages: [
    {
      role: "user",
      content: [{ type: "text", text: "A chubby orange cat napping on a windowsill, watercolor style, warm tones" }],
    },
  ],
});

// Final image, URL valid for 12 hours
console.log(res.choices[0].delta.image.url);

Access Methods

• JS SDK — Node.js server side
• Hy-Image-3.5-preview — full protocol and parameter reference
• Legacy Models — deprecated HY-Image-3.0 usage