TL;DR
- Auto Links = a rule you write. "Whenever the phrase X appears in any post, link it to URL Y." Mechanical. Predictable. You're in charge. Lives at Linkilo → Linking → Add New Auto Links.
- AI Suggestions = Linkilo's brain. Reads each post, decides which other posts on your site it's topically related to, and proposes a link with an anchor picked from the sentence and a short reason. Smart. Contextual. You approve each one (unless you turn on Auto-apply high-confidence). Lives at Linkilo → Linking → AI Suggestions.
Most sites end up using both. They solve different problems.
When Auto Links is the right tool
Pick Auto Links when you already know exactly which phrase should link to which URL.
- A revenue page that needs every mention of a product name to link to it. You sell "the Linkilo Pro license" and want every blog mention of "Linkilo Pro" to link to your pricing page.
- An author bio you reuse. "Written by Jane Doe" should always link to Jane's author page, in every post.
- A campaign URL on a seasonal anchor. "Black Friday deal" → your Black Friday landing page, with a start/end date so it auto-expires.
- A canonical anchor → URL mapping you're enforcing across the site. You decided "espresso machine" links to
/best-espresso-machines/everywhere; Auto Links is the enforcement tool.
Auto Links runs without any AI cost. No provider key needed. The rule fires on every post matching its conditions, and in 3.0 you can click Preview impact on a new rule, or tick rules and use Preview Selected before Create Links, to see exactly which posts would change.
When AI Suggestions is the right tool
Pick AI Suggestions when you want Linkilo to discover opportunities you wouldn't have thought of.
- Filling out internal links on existing content. You've got 500 posts and don't remember which ones cover what. AI scans them and surfaces relevant connections.
- New posts that need contextual links. You just wrote a post; with Analyze on publish on, AI reads it and proposes links to related existing posts based on semantic match — not exact keyword match.
- Cluster building. AI groups posts into topic clusters and favours links between them.
- Catching topically related content you didn't realise was related. A coffee post linking to a productivity post because both discuss "morning routine" — Linkilo's embedding model catches that; an Auto Link rule never would.
AI Suggestions uses your AI provider's credits (OpenAI by default). You pay the provider, not Linkilo. The AI Usage Meter in Settings → AI Settings shows estimated provider spend, and the monthly caps (1,000 requests and $25 by default) stop surprises.
Side-by-side comparison
| Auto Links | AI Suggestions | |
|---|---|---|
| Cost | Free | Your provider's usage charges, capped monthly |
| Setup | Write one rule per anchor → URL pair | Connect a provider key once, process your content, then Build Queue |
| Speed of insertion | Immediate, on every matching post | One capped scan per run; suggestions reviewed in batches |
| Context-awareness | None — it's a literal phrase match | High — uses meaning, not just keywords |
| Best for | Known anchor → URL mappings | Finding new internal-link opportunities |
| Risk | Over-linking if your rule is too broad | Off-topic suggestions if Minimum Confidence Level is low |
How they interact
The two engines run independently, but they respect each other:
- Every AI batch checks the post's existing links before suggesting, so a phrase an Auto Links rule has already linked isn't proposed again for the same target.
- The filters in Settings → Content & Ignored Words apply to every suggestion, AI and legacy alike. Posts in the Skip auto-link & URL changes list are never edited by Auto Links either. Each Auto Links rule also has its own category limits and post-type whitelist.
- Both end up in your post content, so after the next Perform Scan the Anchor Report, Link Cannibalization and Click Analytics cover links from either source.
Common mistakes
Using Auto Links when you should be using AI Suggestions
Symptom: the same exact-match anchor pointing to the same URL in 60 different posts because your Auto Links rule has no per-post cap.
Fix: turn on the one-link-per-post option in the rule's Settings panel. Better: delete the rule and let AI Suggestions propose context-appropriate links with varied anchors.
Using AI Suggestions when you should be using Auto Links
Symptom: AI doesn't reliably link "Free SEO Course" to your course landing page because the post text uses paraphrases like "free course" or "SEO training" and the AI scores them differently.
Fix: add an Auto Links rule for the exact phrases you care about. Use AI for everything else.
Running both with overlapping rules and no cap
Symptom: a post ends up with 12 links to the same target URL.
Fix: Auto Links → one link per post ON. AI Suggestions → Max Suggestions per Post in Processing Settings capped, and Max internal links per post set in Settings → General. The Same URL card on the Summary page and the Link Cannibalization Report flag accidental dupes; Fix Link Health can clean them up in batches.
Suggested setup for most sites
- Start with AI Suggestions to get a baseline of contextual internal linking across all your existing content. Build Recommended Queue is a good first run.
- Identify 5–15 anchor → URL pairs that must always link consistently (product names, revenue pages, author bios).
- Add those as Auto Links rules with one link per post, and check them with Preview impact.
- Let AI Suggestions handle everything else, with Smart autopilot on for ongoing maintenance.
See also
- AI Suggestions
- How to Use the Add New Auto Links Feature
- Is the Auto-Linking Tool Fully Automated?
- How to Improve Linkilo's Link Suggestions
- Link Priorities — which pages deserve either kind of link next
Last reviewed for Linkilo 3.0.0 (October 2026).