Create a scoped token
Enable only AI crawler signals on a MentionLeads API token you can revoke at any time.
MentionLeads AI Visibility tracks server-side requests from ChatGPT, Claude, Perplexity, search indexes, and training crawlers. See which pages they requested and why they came.
Live assistant fetchesA person may be waiting for an answer right now.
Search discoveryYour page is being recorded so it can be cited later.
Model collectionSeparate training crawlers from higher-intent requests.
A small server-side companion catches the requests browser analytics misses, identifies the crawler, removes query strings, and aggregates the useful signal by page and day.
Enable only AI crawler signals on a MentionLeads API token you can revoke at any time.
Detect known crawler user-agents and send their user-agent plus requested URL in the background.
Use waitUntil or a fire-and-forget call so the tracking request never blocks the crawler or your page.
Compare answers, indexing, and training by bot and requested path inside Growth Analytics.
Separate assistant answer fetches from search indexing and model-training collection because they mean different things.
Recognize OpenAI, Anthropic, Perplexity, Google, Bing, Apple, Meta, Common Crawl, and other known user-agents.
See which pricing, alternative, guide, and product pages AI systems request most often.
Keep browser visitors and server-side machine traffic separate, then compare both in the same analytics workspace.
One scoped key can report crawler signals across multiple hosts, with www normalized into the same site.
Use ready-to-adapt examples for Next.js, Express, Cloudflare Workers, curl, and any backend that can send an HTTP request.
Designed for trust
A crawler request shows that an AI system asked your server for a page. That is valuable evidence, but it is not the same as a verified mention in a generated answer. MentionLeads keeps that distinction explicit.
No vague AI promises. These are the practical boundaries of the feature.
AI crawlers usually request raw HTML directly from your server and do not execute your browser JavaScript. MentionLeads therefore uses a server-side middleware or backend call to record the request.
Answer crawlers fetch a page while an assistant may be responding to a user. Indexing crawlers record pages so they can be found or cited later. Training crawlers collect content for model development. MentionLeads reports them separately.
No. It proves that a request with a recognized crawler user-agent reached your page, not that the resulting answer mentioned or cited you. User-agents can also be spoofed, so the dashboard presents this as a visibility signal.
MentionLeads keeps the normalized host and path, identified bot, purpose, and aggregated daily count. Query strings, request bodies, visitor IP addresses, and cookies are not stored for crawler tracking.
Server-side AI crawler tracking is included in Pro and Agency and appears inside Growth Analytics. New accounts receive three days of full Pro without a card.
Start with your product. MentionLeads will help you find demand, build authority, and see what happens next.
Start free