LLM Optimization Score
Ten weighted rules score every post on how quotable it is for a language model.
ChatGPT, Perplexity, Claude and Gemini answer your customers before Google does. WpOptima scores every post against how those models actually read a page — and tells you exactly what to fix.
One-time payment · Unlimited sites · 30-day money-back guarantee
Already running on 100+ WordPress sites — agencies, shops and publishers optimising for AI answers.
5
Scoring dimensions: LLM, GEO, AIO, AEO and Product
8
Schema JSON-LD types generated automatically
24
AI crawlers you allow or block, one by one
0
External API calls needed for scoring
Nineteen modules that read your content the way a language model does, then hand you the fix — not just a number.
Ten weighted rules score every post on how quotable it is for a language model.
Generative Engine Optimization: how well the page survives being summarised.
Measures readiness for AI Overviews and assistant-generated summaries.
Answer Engine Optimization — is the page a direct answer to a real question?
Eight JSON-LD types built and validated automatically: Article, FAQ, HowTo, Product and more.
Links people, places and brands to Wikidata and DBpedia identifiers models already trust.
Rewrites headings into the question form answer engines match against.
Eight assistants powered by Claude, GPT or Gemini — your key, your choice of model.
Scores product pages on the attributes shopping assistants actually compare.
Publishes the manifest assistants read to understand what your site is about.
Allow or block 24 named AI crawlers individually, with Content Signals support.
See which bots visited, when, and which pages they actually read.
A dedicated XML feed that points crawlers at the pages worth citing.
Counts real visits arriving from ChatGPT, Perplexity, Claude and Gemini.
Checks on a schedule whether models mention your brand — and who they cite instead.
Exposes your site to Claude Desktop, Cursor and Cline over Model Context Protocol.
Audits author profiles against E-E-A-T and fills the gaps models look for.
One screen telling you whether the whole setup is actually working.
Panels inside Gutenberg, Classic Editor, Elementor and Breakdance — same scores everywhere.
Ranking first on Google no longer means being the answer. These are two different jobs.
No configuration marathon, no external account, no API key to get started.
Upload the plugin, activate it, paste your key. The setup wizard picks your post types and turns on llms.txt.
Every post gets scored on save. The dashboard sorts your library worst-first, so you know where to start.
Apply the suggestions in your editor of choice, regenerate the schema, and watch the citation tracker.
Not another SEO plugin with an AI label glued on the box.
Scoring, schema and entity work all run on your server. Nothing is uploaded anywhere to produce a score.
Rescoring runs in batches of ten. Thousands of posts will not take your site down.
A single lifetime licence covers an unlimited number of installs. No per-site maths, no renewals.
Gutenberg, Classic, Elementor, Breakdance and Oxygen content is normalised before scoring, so the numbers compare.
Each rule explains what is missing and shows the rewrite, instead of leaving you with a red badge.
Claude, GPT or Gemini — pick a provider, paste your key, keep control of the cost.
No tiers to compare, no seat counting, no renewal email in twelve months.
$75
one-time payment, no renewals
One-time payment · VAT included where it applies · invoice on request
30-day money-back guarantee, no questions asked. Your key arrives by email within minutes and activates on every site you run.
The free tier is not a trial. It keeps working, with no key and no time limit.
No. All five scoring engines, the schema generator and the entity work run on your own server. Content only leaves your site if you deliberately switch on the AI content tools and supply your own Claude, OpenAI or Gemini key — and then it goes to that provider, not to us.
As many as you want. One key, unlimited installs, including client sites if you are an agency. Activate it in Settings → License on each install; deactivating frees nothing up, because nothing is counted.
No. You pay $75 once and keep the plugin and its updates for good. There is no renewal and no expiry date on the key.
No. WpOptima works on a different layer — answer quality, structured data for models, and AI crawler control. Keep your SEO plugin for titles, sitemaps and redirects.
Gutenberg, Classic Editor, Elementor and Breakdance all get a WpOptima panel, and Oxygen content is scored correctly too. Builder markup is normalised first, so the same text gets the same score regardless of how it was built.
Not for scoring, schema, entities or crawler control — those need no key at all. A key is only required for the optional AI writing assistants.
It writes per-bot rules into robots.txt for 24 named AI crawlers, adds Content Signals, and logs which of them really visited your site and when. You decide individually who may train on your content and who may cite it.
Yes. Product AI Score rates product pages on the attributes shopping assistants compare, and Product schema is generated automatically.
Write to ai@wpoptima.com within 30 days and you get your money back. No forms, no interrogation.
Allow the ones that send traffic, block the ones that only take. One switch each.
GPTBot
OpenAI
OAI-SearchBot
OpenAI
ChatGPT-User
OpenAI
ClaudeBot
Anthropic
Claude-Web
Anthropic
anthropic-ai
Anthropic
Google-Extended
GoogleOther
PerplexityBot
Perplexity
Perplexity-User
Perplexity
Bingbot
Microsoft
MSN-Copilot
Microsoft
Meta-ExternalAgent
Meta
FacebookBot
Meta
xAI-Bot
xAI
Bytespider
ByteDance
Amazonbot
Amazon
Applebot-Extended
Apple
cohere-ai
Cohere
MistralAI-User
Mistral
DeepSeekBot
DeepSeek
CCBot
Common Crawl
AI2Bot
Allen Institute
PetalBot
Huawei
One key, every site you own, updates for life. If it does not earn its price back, you have 30 days to ask for a refund.
Large language models do not rank pages, they read them and decide what is worth repeating. WpOptima scores each post against the patterns those models reward — a definition in the opening paragraph, concrete numbers, short paragraphs, clean lists and headings phrased as questions — then shows you the specific line to change.
Generative Engine Optimization asks whether your page survives being summarised, and AI Overview readiness asks whether an assistant can lift a clean answer from it. WpOptima measures both, generates the JSON-LD that backs the claim, and links your entities to Wikidata identifiers models already recognise.
Answer engines return one response, not ten blue links. The AEO score checks whether a page answers a real question directly, whether the answer is findable without scrolling, and whether the structured data confirms it — so that when ChatGPT or Perplexity replies, it is your site in the citation.