The AI citation-tracker MCP + API: are your URLs getting cited by AI? Citations are the new backlinks.
Google's backlink graph is fading; AI citation graphs are forming. Pass any URL to /v1/citations, get back the prompts that surface it, the models that cite it, and the frequency. Decide which pages to refresh, which to archive, and where to invest in new content.
- $1 free on signup
- Credits never expire
- No subscription
- MCP native
1. You ask your AI, in plain English Is penflow.com/blog/seo-vs-aeo getting cited by AI? 2. Cite42 MCP checks every model for you → ChatGPT · Claude · Perplexity · Gemini 3. You get back, in plain English Cited 4 times across 3 prompts — strongest on Perplexity.
Ask in plain language. Get structured JSON.
Each is one call against the endpoint, ask in plain English through the MCP, or hit the API directly. Same key, same credit balance.
{ "cited": true, "prompts": 3, "citations": 4 }{ "by_model": { "chatgpt": 3, "perplexity": 4 } }{ "gap_prompts": ["best ai blog tool", "seo vs aeo"] }Backlinks measured the last web. AI citations measure this one.
A backlink tool tells you who linked to a page. It says nothing about whether ChatGPT, Claude, Perplexity, or Gemini treats that page as a source worth quoting. Those are two different graphs, and only one of them is growing. Pass any URL to /v1/citations and get back the prompts that surface it, the models that cite it, and how often, in one JSON shape.
A page can rack up backlinks and still be invisible inside AI answers. The reverse is true too: a thin post can get cited four times on Perplexity for the exact category prompts your buyers ask. /v1/citations checks all four models at once so you see the citation graph that actually feeds AI answers, not a proxy from the Google era.
- See per-model citation counts, not a single blended score.
- Diff a URL against a competitor's to find the prompts where they get cited and you don't.
- Re-run the same URL after a refresh and measure the citation delta.
Every model, one API key.
One call reaches all 4 AI models, ChatGPT, Claude, Perplexity, and Gemini, or pass “models” to narrow to the ones you need. Billed per model queried.
- Web search on
ChatGPT
OpenAI
The highest-traffic AI surface
- Web search tool
Claude
Anthropic
Cited, source-grounded answers
- Web-grounded
Perplexity
Perplexity Sonar
Always answers from live sources
- Google Search grounding
Gemini
Google
Google's own model layer
- Live SERP
Google AI Overviews
Google Search
The AI summary atop Google results
Real scenarios. Real numbers.
Real workflows, real call counts, real totals, see what this looks like before you commit. Credits never expire.
- content lead
Audit a content library
Check 50 of your top blog URLs across all four models to see which earn AI citations and which are dead weight. 50 calls · $10.00 total.
- AEO consultant
Prove citation lift to a client
Baseline 40 client URLs, refresh the content, re-run the same 40 a month later, and show the citation delta. 80 calls · $16.00 across the engagement.
- competitive analyst
Find a competitor’s citation gaps
Run your URLs and a competitor’s through the same prompt set, diff the results, and target the prompts where they win and you don’t. 60 calls · $12.00 total.
Track AI citations, not the inbound links that no longer move AI rankings.
Ahrefs and Semrush tell you who linked to a page. Useful for the Google-era ranking graph, useless for knowing whether a model considers your URL a credible source. Citations are the inbound reference that AI answers are actually built from. Same question, are people finding this page, two different graphs.
| Cite42 | A backlink tool | |
|---|---|---|
| What it tracks | Whether AI models cite the URL as a source | Whether other websites link to the URL |
| Coverage | ChatGPT, Claude, Perplexity, and Gemini in one call | A crawl of the public web's link graph |
| Real-time | Every call runs live against current model APIs | Periodic crawl; the index lags reality by days or weeks |
| Integration | One endpoint or MCP tool, one JSON shape, callable by an agent | A dashboard built for humans, export and reconcile by hand |
| Repeat checks | Identical URL checks replay from a cross-customer cache at $0 | Seat-based monthly subscription whether you check or not |
| Cost, 4-model check | $0.20 ($0.05 per model), or $0 on a cache hit | A fixed monthly plan, billed even on idle months |
Pay per call. $25 minimum deposit. Credits never expire.
This page's endpoints below. See /pricing for the full per-endpoint table.
| Endpoint | Price | What you get |
|---|---|---|
| POST/v1/citations | $0.05/model | Find prompts and models that cite a given URL, $0.05 per model per call, all four for $0.20 |
Related endpoints
Same API key, same credit balance, different question.
Questions about ai citations.
See which URLs AI is citing today.
$1 free on signup, covers your first few citation checks.