Every SEO tool claims to be "AI-powered." SEObolt uses AI in two distinct, clearly separated ways — and we are upfront about what is AI, what is rule-based, and what we don't do at all.
The "AI SEO" category has become a marketing swamp. Half the tools bolt a chatbot onto a rank tracker and call it AI. Others use the word for anything with an algorithm. SEObolt splits the picture into two precise layers so you always know what is actually happening.
The first layer is AI we call — external models (OpenAI, Anthropic, DeepSeek, Google) that generate copy on your behalf inside WordPress. You trigger it, you see every word before it saves, and you can switch providers at any time.
The second layer is AI we track — external models (ChatGPT, Claude, Perplexity, Gemini, DeepSeek) whose answers we monitor on the SEObolt SaaS dashboard so you can see whether your brand is being cited in AI search results.
Everything else in SEObolt — the 57-point SEO Score, the 11-factor Schema Score, the 10-point Local SEO Score, the 37-factor GEO Score — is rule-based. Not AI. Deterministic, explainable, repeatable math over your content and your analytics data.
Side by side, the distinction is easy to spot. One layer writes for you. The other watches whether AI writes about you.
Content AI lives in the WordPress post editor. You click a button, the plugin sends a tightly-scoped prompt (your current content, focus keywords, and SEO constraints) to the provider you picked, and you get drafts back. Nothing is auto-published, and nothing leaves your site until you press the button.
Multiple SEO-tuned title variants constrained to the 60-character SERP sweet spot, each grounded in your focus keywords.
Click-worthy meta descriptions at the 155-character limit, with natural keyword placement and search-intent alignment.
Specific, actionable changes to lift your on-page SEO score — headings, sections, examples, and gaps against your target keyword.
The AI scans your existing content library and surfaces contextually relevant internal-link candidates for your current post.
Related keywords, long-tail variations, and semantic matches for your target topic — scoped to your default country and audience.
Expand a thin heading or paragraph into fully-developed prose with examples and supporting detail — without drifting off your keyword.
Descriptive, accessibility-friendly alt text for images, grounded in your post context so it reads like a human wrote it.
Auto-draft FAQ sections from your content, ready for FAQPage schema and eligible for rich results on Google.
BYOK or credits. Bring your own API key from OpenAI, Anthropic, DeepSeek, or Google and pay the provider directly with zero SEObolt credits consumed. Or use the monthly credit allocation included with your SEObolt tier — no separate account required.
LLM Rankings runs on the SEObolt.io SaaS dashboard. It queries five major LLMs on the prompts your customers actually ask, records whether your brand is cited, and compares your share of AI citations against competitors.
This is not something the plugin does locally — it requires a cloud backend to run prompts on a schedule, store historical data, and surface trends. Available on Business and Agency plans; each keyword check uses 1 AI credit (covering all providers), or bring your own LLM API keys to bypass credits.
For every tracked prompt, see which LLMs mentioned your brand, which linked to your pages, and which sent customers to a competitor.
Your AI citation share vs rivals across all five LLMs and all tracked prompts, trended over time.
Get notified when you gain or lose citations on specific prompts — the closest thing to "rank tracking for AI search."
Auto-build an llms.txt manifest from your GSC-indexed pages to help AI crawlers discover your most important URLs.
GEO Score is a 37-factor, rule-based scoring engine that grades how well a page is optimized for AI citations. It's built on peer-reviewed research from Princeton, Georgia Tech, and UC Berkeley — not on an LLM. It tells you what to fix on your content; LLM Rankings then tells you if it worked.
The loop:
Because GEO Score is rule-based, its output is fully explainable — every point deduction comes with a specific factor and a specific fix. No LLM hallucinations, no "trust me, the AI said so."
To remove any ambiguity, here's the exact line between AI and rule-based features inside SEObolt.
| Feature | Type | Details |
|---|---|---|
| Content AI (titles, meta, suggestions, link hints) | AI (we call) | OpenAI, Claude, DeepSeek, or Gemini — your pick |
| LLM Rankings (citation tracking) | AI (we track) | Queries ChatGPT, Claude, Perplexity, Gemini, DeepSeek |
| 57-point SEO Score | Rule-based | Deterministic checks on HTML, headings, links, meta |
| 11-factor Schema Score | Rule-based | JSON-LD validation + completeness checks |
| 10-point Local SEO Score | Rule-based | NAP consistency, hours, LocalBusiness schema |
| 37-factor GEO Score | Rule-based | Peer-reviewed GEO research — not an LLM |
| Content decay alerts, keyword position tracking | Rule-based | Threshold logic over GSC and SERP data |
SEObolt is not an AI content writer, not an automated blog generator, and not a backlink farm. If you're looking for a tool that writes entire posts and auto-publishes them, SEObolt is the wrong product. It's a rank-tracking and optimization platform that uses AI where AI is genuinely useful, and uses deterministic math everywhere else.
See the two AI layers and the rule-based scoring engines working together on a single license. No marketing swamp, no "AI everywhere" handwaving.