本工作流为 Ideogram4 文生图工作流。Ideogram4 模型的优点是可以精准布置画面元素的位置。

V1 版本为 2 步快速出图模板；

V2 版本为增加 Krea2 修复 Ideogram4 出现的画面崩坏、人体畸形；

V3 版本为增加自动提示词处理。

使用方法：

1、在文本框内填写相应的提示词。

2、在图形设置区添加对象，设置对象的文本描述、位置、大小和颜色。

3、如果有参考图，可以将参考图扔到工作流界面，然后使用图形设置区的取色器选取颜色。

使用的加速lora网址：

https://civarchive.com/models/2711950/ideogram-4-turbotime?modelVersionId=3046680

This workflow is a text-to-image pipeline built for Ideogram 4. A key advantage of the Ideogram 4 model is its precise control over the placement of visual elements within a frame.

Version V1: A 2-step fast image generation template.

Version V2: Integrated with Krea2 to fix broken or distorted artifacts produced by Ideogram 4.

Version V3: Added automatic prompt processing functionality.

How to Use

Enter your corresponding prompts in the text input box.

Add objs in Graphic Settings, adjust their prompt, pos, size & color.

If you have reference images, drag and drop them directly onto the workflow canvas. You can then use the eyedropper tool in the graphic settings panel to sample colors.

Accelerated LoRA Used

Download link for the speed-up LoRA:

https://civarchive.com/models/2711950/ideogram-4-turbotime?modelVersionId=3046680