Deactivate Free version before activating the Pro version. Your settings will be preserved.
Install the Free version of Paca AI from the WordPress Plugins Marketplace, or get Pro version and upload the ZIP file via the WordPress Plugins section.
After activating the plugin, navigate to the plugin settings page and connect your AI API key to begin using AI features.
The settings and features described in this documentation apply to Pro versions of the plugin, unless otherwise noted.
Enable AI features for all post types, or select specific types as needed.
This feature is optional; "post" is selected by default.
Set the maximum length (in words) for AI-generated excerpts.
The free version defaults to 25 words, while the Pro version defaults to 50 words.
Create a custom prompt to be used when generating excerpts.
In the backend, the prompt will be combined with the post content.
Allow the AI to utilize custom fields attached to posts for improved results.
Adjust the size of AI-generated images.
Available options are 1024x1024, 1024x1792, and 1792x1024. The default is 1024x1024.
This feature is available only when OpenAI models are enabled.
Create a custom prompt to be used when generating images.
In the backend, the prompt will be combined with the post content.
This feature is available only when OpenAI models are enabled.
If the AI has generated an excerpt, it will be used to create images, resulting in smaller and more cost-effective requests.
This feature is available only when OpenAI models are enabled.
Enable the AI to analyze your site and generate a summary based on the latest content.
Choose how frequently the AI should perform this analysis (daily, weekly, monthly, etc.), and specify the number of posts to include per post type.
This process runs automatically in the background at your selected intervals. Once complete, the AI-generated summary will be available for review.
Specify a selector for the main content area from which the AI will extract content.
The default selectors are: main, article, [role="main"], .entry-content, .post-content, .content, #content, .main, or #main
If your page's main content does not reside within the default HTML elements, you can specify a custom selector for the AI to use.
For example: .my-main-content or #my-main-content
Specify a selector where the AI will place the generated summary content.
By default, a modal window will appear.
To use a custom area, add a selector where the AI should attach the generated summary.
For example: .my-summary or #my-summary
Set the cooldown period (in seconds) before another AI request can be made.
By default, the cooldown is 10 minutes before another request can be made.
Decreasing this limit may significantly increase the cost per visitor.
The summary is cached in localStorage for 24 hours, so users will see the same summary during that period.
For private browsing, the summary is cached in memory; however, refreshing the page will clear it.
The cooldown helps prevent request abuse and keeps your costs lower.
Set the maximum word count for the AI-generated summary in the frontend widget.
Default: 60.
Add the .paca-summary-trigger class to an HTML element.
When this element is clicked, the AI will generate a summary based on the page content and display it in a modal window, or attach it to a Frontend Widget Selector for Summary Content if specified in the settings.
Example: <button class='paca-summary-trigger'>Summarize Page</button>
Add [paca_summary_button text='My Button' class='my-button'] shortcode anywhere in your post editor.
Optional: text attribute allows you to customize the button text, and class attribute lets you add custom classes for styling.
Allows developers to customize the maximum number of summary requests a user can make within a defined time window.
This helps prevent AI API abuse when using frontend widget, especially in private browsing.
Example: add_filter( 'paca_summarize_max_requests', 5 );
Default: 5 requests.
Lets developers set the duration (in seconds) for the rate-limiting window during which the request count is tracked.
This helps prevent AI API abuse when using frontend widget, especially in private browsing.
Example: add_filter( 'paca_summarize_time_window', 15 * MINUTE_IN_SECONDS );
Default: 15 minutes.