A report on "How to Add Internal Links to AI-Generated Articles"

The graded article is 906 words, written by DURBUN from one keyword. Open it below to read what the models read. The report took 14.4 seconds and cost 534 credits.

Scores by axis

Each axis is the average of the three models. Where they disagreed the spread is shown, because a wide spread usually means the article is ambiguous rather than bad.

Search intent

5.7

Writing quality

6.3

Topical coverage

5.7

Information gain

5

Evidence and credibility

6

Structure

5.7

Spam safety

6.3

Strongest: Spam safety, Writing quality. Weakest: Information gain, Search intent, Topical coverage.

What each model said

GPT-5.6 Luna, 6/10. The article offers a reasonably clear DURBUN workflow and includes useful operational details about link caps, supported workflows, and review checks. However, the target query asks how to add internal links to AI-generated articles generally, while most of the page is a product-specific explanation with limited guidance for other AI tools or a normal WordPress workflow. It is useful documentation-style content, but not strong enough to outrank a comprehensive, platform-neutral how-to.

GPT-5 Nano, 7/10. Solid how-to guide that explains a concrete workflow for adding internal links to AI-generated content using a specific tool (DURBUN). It is clear and actionable, but it is tightly coupled to the product and lacks broader perspectives or independent validation. It would rank well for the exact process but may lower credibility for non-users or competitors.

Gemini 3.8 Flash, 3/10. This article hijacks a broad, platform-agnostic informational query and turns it into product documentation and sales copy for a specific tool called DURBUN. A searcher looking to solve this problem needs to understand general methodologies—such as prompt-based URL feeding, dedicated plugins like Link Whisper, or vector/embedding-based post-processing—rather than reading pricing tables and workflow logs for an unintroduced proprietary SaaS.

What it got right

• “A useful destination should give the reader information that the current paragraph does not fully provide, rather than appearing only because it shares a keyword.” This gives a sound relevance rule that is more useful than simply recommending keyword matching.
• “Concise anchor text should tell readers what they will find, such as email newsletter examples or technical SEO checklist, rather than using ‘click here.’” This provides concrete anchor-text guidance and an example.
• “Look for repeated anchors, duplicate destinations, broken links, and links that make the paragraph harder to follow.” This is a practical pre-publication quality-control checklist.
• “Auto places up to 6 links, or up to 8 in articles of 2,000 words or more.” This is a specific product behavior that makes the DURBUN instructions actionable.
• "Start with the posts already available on the site."
• "Concise anchor text should tell readers what they will find, such as email newsletter examples or technical SEO checklist, rather than using “click here.”"
• "Auto places up to 6 links, or up to 8 in articles of 2,000 words or more."
• "Review the links in the finished article and keep the ones that lead to useful next information."
• The core editorial advice regarding anchor text and contextual placement is sound: "Concise anchor text should tell readers what they will find, such as email newsletter examples or technical SEO checklist, rather than using 'click here.' Skip image captions and never link an article to itself."

What it wants fixed

• The page answers a general how-to query mainly through product promotion: “DURBUN adds internal links to AI-generated WordPress articles...” and “DURBUN does this matching from paragraph content and existing post titles.” Replace the opening with a platform-neutral step-by-step method, then present DURBUN as one implementation option.
• It does not adequately explain how to perform the task without DURBUN. The only manual guidance is the short FAQ instruction, “Open the article in the WordPress editor, select the words that describe the destination, choose the link option, and enter the post URL.” Add a complete manual workflow covering content inventory, topical mapping, anchor selection, insertion, validation, and a reasonable link-density decision.
• The section “Select the Link Count for the Publishing Run” is more about DURBUN billing and campaign coverage than how many internal links an article should have. The table repeats “No credits” and “1 credit per output word” without explaining editorial criteria. Replace or supplement it with guidance based on article length, topic complexity, page importance, and reader usefulness, while keeping billing in a brief product note.
• “Start with one article and compare the links DURBUN adds...” is too thin to justify its own section and reads like a generic mini-conclusion. Fold it into the review workflow and provide a concrete test, such as checking destination relevance, anchor diversity, crawlability, and whether links support the article’s intended conversion path.
• The FAQ includes “What can I use as the starting point for an internal-linking workflow?” and answers with “A WordPress post list, a maintained content inventory, or your site’s sitemap...” but does not explain how to turn that inventory into a topical map or identify orphan pages. Add those steps or remove the FAQ so the page’s structure is less templated.
• "DURBUN reads the article's paragraphs beside the titles of posts from the connected WordPress site or the website address you enter." This ties the guidance to a specific tool, potentially limiting applicability without mention of alternatives.
• "Read where the links landed before publishing. Connect the site, turn on inline linking with Auto, and review the finished article before it goes live." This suggests a workflow that may assume certain features exist, with no independent guidance for readers without the tool.
• The article is heavily product-centric and lacks discussion of generic best practices outside the DURBUN ecosystem, such as performance considerations, accessibility implications of internal links, or how to measure impact beyond basic warnings."
• The article immediately hijacks a generic informational search into product documentation: "DURBUN adds internal links to AI-generated WordPress articles by matching each paragraph with relevant posts already on your site." Replace this tool-specific pitch with an objective overview of how internal linking works in AI workflows across prompts, plugins, APIs, and site architectures.
• Major technical and standard workflows are completely missing from topical coverage. An authoritative guide on this topic must cover prompt injection (passing sitemaps to LLMs), dedicated CMS plugins (e.g., Link Whisper, Rank Math), and programmatic embedding/vector search approaches.
• A commercial pricing and credit table is forced into an informational query guide: "| Publishing Workflow | Article Coverage | Inline-Link Cost | Article Billing | ... 1 credit per output word". Delete this software documentation table entirely and replace it with a comparison of internal linking methods.
• The FAQ includes filler entries that feel tacked on solely for search engine scaffolding rather than answering hard implementation challenges: "### Can I add internal links manually in WordPress?" Remove elementary filler and address actual technical pain points, such as hallucinated URLs or managing anchor text diversity at scale.

How to read the score

The grader keeps nine and above for pages it would rank above the current results and publish itself, and it puts most published writing on the web at six or seven. The verdict follows the average: 8.5 and above reads Would outrank the current results, 7 Competitive, worth publishing, 5.5 Publishable after the fixes below, and anything lower Not ready to publish. Every fix above is quoted from a model, so each one points at a passage you can change.