Automate Internal Linking: Free AI Plugin for WordPress SEO

Managing internal links is a cornerstone of effective SEO, yet it often becomes a tedious, manual task. A new, free WordPress plugin leverages artificial intelligence to automate this process directly within your content editor. This tool uses vector search technology to suggest semantically relevant internal links, transforming a time-consuming chore into an efficient, integrated workflow. Built with popular AI APIs, it brings advanced semantic understanding to any WordPress site.

The journey from conceptual Python scripts to a practical plugin began with a clear goal: to eliminate the friction of switching between tools. Most content creators and SEO professionals live inside their WordPress dashboard, not a command line. This solution bridges that gap by embedding powerful AI-driven linking capabilities directly into the familiar editing environment. The plugin is a functional proof-of-concept, built using Google Vertex AI, OpenAI, and Pinecone to manage links intelligently.

Getting started requires a few setup steps, but no deep technical expertise is needed. First, you must create a Pinecone vector database. Registration is straightforward on their website. You’ll need to create an index with a dimensionality of 768, and note your API key and database host for later configuration.

Next, an OpenAI API key is required. Log into your account, ensure you have a small credit balance (as little as $5) attached for billing, and generate a new secret key. Remember to copy it immediately, as you won’t be able to view it again. The cost for this service in the plugin is minimal, typically around fifty cents per month.

Finally, you need a Google Service Account JSON key. This involves visiting the Google Cloud Console, enabling the Vertex AI API, and setting up billing, even for the free tier. Once enabled, you can create a service account and download the necessary JSON key file. The monthly cost for using Vertex AI through this plugin is similarly low, around thirty cents.

With all keys gathered, you can install the plugin ZIP file on your WordPress site, always recommended on a staging site first. The settings panel allows you to input your API credentials and select which post types to index, giving you granular control over your content.

Before the plugin can suggest links, your existing content must be indexed into the vector database. This automated process replaces manual Python scripts. The plugin scans your published posts and pages, extracting key content like the title, meta description, excerpt, opening paragraph, and H2 headings with their subsequent text. This composed copy is converted into a 768-dimensional vector via an embedding model and stored in Pinecone alongside the post ID. This focus on primary topics prevents the vectors from being diluted by less relevant content. Indexing is efficient; for example, 25,000 articles can be processed in about 50 minutes. While this initial indexing is a one-time task with a small cost (roughly $1-$2 for tens of thousands of articles), the plugin automatically updates, adds, or deletes vectors whenever you publish or modify a post, keeping your index current.

The core functionality shines when adding links. The plugin operates in two primary modes: suggesting links for a manually selected anchor phrase or automatically parsing an entire article to generate bulk suggestions. It is fully compatible with both the Gutenberg and Classic WordPress editors.

Within the Gutenberg editor, a dedicated sidebar panel provides automatic internal link suggestions. You can configure filters, such as limiting suggestions to a specific category like “Analytics” or to articles published within a certain timeframe, which acts as a retrieval-augmented generation (RAG) technique to boost accuracy. Clicking “Bulk Auto Suggestion” triggers a sophisticated process: the plugin uses OpenAI’s model to extract key entities and phrases from your draft, converts those phrases into vectors using Google Vertex AI, and then queries Pinecone to find the most semantically relevant existing articles. The results appear in the sidebar, where a single click applies the link.

The Classic editor experience is equally streamlined. A “Suggest Links” dropdown is added to the toolbar. You can select any text phrase and choose “Suggest by Anchor Text,” or run the same bulk suggestion process to get AI-powered recommendations throughout the content.

While several other internal linking tools exist, both free and paid, this custom solution was built for a specific reason: granular control. Existing plugins often lacked the flexibility and fine-tuning capabilities required for advanced semantic linking strategies. Building from scratch ensured complete independence from third-party platforms and total control over the AI models and logic. The only costs are the minimal, predictable fees from the API providers, and it’s a best practice to set budget alerts to manage these.

This project demonstrates a practical application of AI and vector databases for SEO. The shared plugin files serve as an educational foundation, illustrating how to move from theoretical embeddings to a functional tool. It is provided without official support but as a starting point for the community. All data processed remains under the site owner’s full control, with no external collection or sharing. For those interested in extending its functionality or discussing the code, the creator is available for questions.

(Source: Search Engine Journal)