How do you create a full set of Xiaohongshu visual posts?

First, organize your content in Markdown and send it to the 文卡网 AI assistant. Choose the “Xiaohongshu 3:4” or “3:5” aspect ratio, then add this instruction:

Create a 3–5-page Xiaohongshu carousel, with one key point per page.

The 3–5-page range is just an example. Adjust it to fit your content. AI will split the content into key points and turn each one into a separate visual card.

Steps to follow:

Organize your content in a Markdown document.
Choose the Xiaohongshu 3:4 or 3:5 aspect ratio.
Add: “Create a 3–5-page Xiaohongshu carousel, with one key point per page.”

How can you use fewer credits?

An AI design conversation starts at 1 credit. Longer conversations use more tokens and may cost more credits. Charges depend on the selected model and actual token usage. Clarification-only turns that produce no design are not charged.

Credits are calculated from the model's token usage. Fewer tokens mean fewer credits. Here are a few ways to reduce token usage:

Plan the length of your content. Aim for enough content to make one set of visual posts, such as a WeChat Official Account article. Very long content gives the model more information to process and increases token usage.
Avoid long conversations in the same project. More conversation turns mean a longer context and more tokens. If you're returning after a while, consider starting a new project.
Describe your reference style in advance. If you have a reference image, you can use another model's web app, such as Doubao or Gemini, to analyze its style and turn it into a design prompt. Send that prompt to the 文卡网 AI assistant. This can substantially reduce token usage.

How do the models differ?

The model selector in the creation field shows the currently available models. Their prices, speed and image-understanding capabilities differ. Before attaching a reference image, check that the selected model supports it.

Try the default model first, then choose a model for your task. Actual charges depend on tokens and the selected model; not every task costs exactly 1 credit.

Give the model one job

The 文卡网 AI assistant's system prompt focuses on layout and design. It can also brainstorm and summarize when asked, but those tasks take more reasoning time and use more tokens.

A better workflow is to organize your content with another model first, then give it to the 文卡网 AI assistant and let it focus on the design. This usually improves the result while saving credits.