Performance

This page covers recommendations for running Supercharger efficiently, especially on larger or higher-traffic WordPress sites.

Choose the Right Provider for Each Task

Supercharger lets you use different providers for different purposes:

• Text generation (summaries, quotes, highlights): Use any supported provider. Lighter models tend to be faster and sufficient for most tasks.
• Embeddings (semantic indexing and recommendations): OpenAI is required, specifically using text-embedding-3-large.

For text generation tasks, there is no need to use the largest or most powerful model available. Most of Supercharger’s generation prompts are well-defined and benefit more from reliability than raw intelligence.

Enable Response Caching

Supercharger can cache AI-generated results so that each post only requires a single API call per generation cycle. Caching is configured in Supercharger → Settings → General.

With caching enabled:

• A post’s summary, key moments, or pull quote is generated once and stored.
• Subsequent page loads serve the cached result instantly — no API call required.
• Cache is invalidated automatically when post content changes.

Caching is strongly recommended for all production sites.

Limit Post Types

Only index and process post types that benefit from AI features. If your site has a large catalog of product pages, author profiles, or tag archive pages registered as post types, exclude them from Supercharger’s processing.

Go to Supercharger → Settings → Content and select only the post types that publish editorial content.

Gradual Indexing on Large Sites

If your site has thousands of posts, the initial indexing will take time. This is expected. WordPress cron processes posts in batches to avoid memory spikes.

If you want indexing to complete faster:

• Set up a real server-side cron job to trigger wp-cron.php every minute (instead of relying on user traffic to trigger it).
• Avoid running full re-indexes unless necessary.

WordPress Cron Configuration

Supercharger’s background processing depends on WordPress cron. For the most reliable results, replace WordPress’s pseudo-cron with a true server cron job:

Disable WP-Cron in wp-config.php:

php
define('DISABLE_WP_CRON', true);

Add a server cron job (example for cPanel or Linux servers):

wget -q -O - https://yoursite.com/wp-cron.php?doing_wp_cron >/dev/null 2>&1

This runs the cron every minute, ensuring indexing and generation tasks are processed promptly.

Monitor API Usage

Keep an eye on your AI provider’s usage dashboard, especially during:

• Initial indexing of a large content library
• After a full re-index operation
• When multiple modules with frequent API calls are active simultaneously

Use the Logs page to see Supercharger’s API activity and identify any unusually expensive operations.

Keep the Index Up to Date

Supercharger automatically re-indexes posts when their content changes. However, if you make bulk edits to many posts at once (e.g., via a plugin or direct database edit), the change detection may be bypassed. In that case, trigger a selective re-index from Supercharger → Tools.