Free tools
Free AI visibility tools
AI visibility is whether assistants like ChatGPT, Claude, Perplexity and Google's AI answers name your brand when someone asks a buying question — and whose pages they cite when they do. It is not the same thing as ranking in search: an assistant reads a handful of sources, summarises them, and names one or two companies, so sitting on page one of Google and being completely absent from the answer happen together all the time.
Two things decide whether you show up. Whether AI crawlers can read your site at all, and whether the pages they can read actually answer the questions your buyers ask. Each tool below checks one part of that, free, in a browser tab.
Four checks, no account
Each one runs on its own and gives you something you can act on the same day. Two of them never ask for an email at all.
AI visibility grader
“How visible is my brand in AI answers right now?”
Grades your site out of 100 from a real AI answer about your brand, with the readiness checks behind the score and a shareable public grade page and badge.
Email requiredOpen the tool →llms.txt generator
“Which pages should AI assistants read first?”
Builds a ready-to-publish llms.txt from your homepage and your own sitemap. Every link comes from a page you already publish — nothing is invented.
No email neededOpen the tool →AI crawler access check
“Can GPTBot and ClaudeBot even read my site?”
Reads your live robots.txt and reports six AI crawlers as allowed, blocked or unknown — plus the exact lines to paste back to unblock the ones you did not mean to shut out.
Email optionalOpen the tool →AI prompt set generator
“What should I be asking ChatGPT about my market?”
Turns a short description of your brand into up to ten buyer questions worth watching, for one buying stage — awareness, evaluation or purchase.
Email requiredOpen the tool →
CiteCue is continuous monitoring and AI auto-fixes.
Everything above tells you where you stand today, once. The problem is that AI answers are not stable: the same question asked next week pulls different sources, names a different competitor, and reflects whatever anyone published in between. A single check cannot tell you whether you are gaining ground or quietly losing it.
So CiteCue does both halves on a schedule: it watches what the assistants say about you as it changes, and it writes the fix for what it finds — published to your site in one click where you have connected a CDN, or handed to you as a draft where you have not.
- Recurring scans across several AI platforms — your prompts asked on a schedule, with live web search, and every answer and cited source recorded.
- Tracked prompts over time — mention rate and brand position per question, scan over scan, so a change has a date and a cause instead of being a hunch.
- Citations and competitor share — which domains the assistants actually trust on your topics, and how much of the answer space each rival holds.
- AI Auto-Fix — a prioritised queue of what to publish or change, with the draft written for you and, where you have connected a CDN, applied to your site in one click.
- Agent-usability testing — an AI agent sent to complete real tasks on your site, reporting exactly where it gets stuck.
How the four fit together
They stack, and the order matters. Start with access: if GPTBot or ClaudeBot is blocked in your robots.txt, nothing else you do can be quoted, because the crawler never sees the page. Then publish a map — an llms.txt points assistants at the pages you most want read, which is a hint rather than a rule, but a free one. Then decide what to measure: the prompt set is the list of buyer questions where an answer names somebody, and it is the thing worth watching over months. The grader is the baseline — one real AI answer about your brand, scored, so you know what you are starting from.
Why a question that names your brand measures nothing
The most common mistake is to ask an assistant "is my brand any good?" and take the friendly answer as evidence. A question that names you guarantees the answer repeats you. What matters is the category question a stranger would type — "best CRM for a small sales team", "reliable dentist near me" — where the assistant picks who to name. That is why the prompt set generator produces awareness and purchase questions that never mention your brand, and keeps the ones that do in their own evaluation stage.
Do these tools change anything on my site?
No. Every tool here reads what is already public — your homepage, your robots.txt, your sitemap — and hands the output back to you to publish yourself. Nothing is written to your site, and the two tools that spend nothing on our side (the llms.txt generator and the crawler access check) do not ask who you are. The grader and the prompt set generator each run a real AI model, which is why they ask for an email in return, and each is limited to a small number of free runs per day.