Citaim

AI visibility tracking · OpenAI web search · Claude · Perplexity · Gemini · Google AI Overviews

Is your brand in the answer?

When buyers ask ChatGPT, Perplexity or Gemini what to choose, Citaim tracks whether they hear your name — every week, against the competitors they hear instead.

Free plan: 3 prompts, one complete baseline, no card.

Acme Robotics
acmerobotics.com · weekly · 3 engines
▲ +5 pts this week
36%
Share of voice
58%
Answers naming you
Positive
Tone of mentions
0%25%50%75%100%4 May18 May1 Jun15 Jun29 Jun6 Jul36%58%
Share of voice Answers naming you
Who wins the answers
  • Roomba27×
  • Roborock19×
  • Acme Robotics16×
  • Ecovacs11×

Sample project. Your own numbers appear after the first cycle.

970
well-known sites scored on the public leaderboard
76/100
median AI & web readiness — most brands have easy wins left
17%
so locked down that AI crawlers can't read them at all

What you see every week

Not one lucky chat — a rate, sampled from clean sessions on every engine, so the numbers are comparable cycle to cycle.

  • Share of voice over time

    One number per cycle — and the trend line that shows whether your GEO work is landing.

  • Who wins the answers

    The brands the AI names instead of you, ranked. Your real competition for the mention.

  • Where AI cites you

    The sources the answers lean on. Get featured there and you climb.

  • Alerts and a weekly digest

    A move of 10 points or a drop-out triggers an email. Monday brings the summary.

How it works

  1. 1
    Add the questions your buyers ask

    Type them, or let the AI suggest ten from your domain. Add the competitors you lose deals to.

  2. 2
    We ask every engine, every cycle

    Each prompt runs several times on each engine from clean, stateless sessions — the answer a stranger gets.

  3. 3
    Watch the trend, fix what moves it

    Share of voice, competitors and cited sources per cycle, with alerts when something changes.

Not ready to sign up? Ask one engine one question.

A free, one-shot AI visibility check: what the model knows about your brand, whether it’s accurate, and the first GEO fixes. No account. Tracking is what turns this snapshot into a trend.

Plans that pay for themselves

Start free. Upgrade when you want more prompts, more engines, competitors and alerts. Annual billing is two months free.

Free

$0no card

See where you stand, every week.

  • $1.00 AI credit, no card
  • 3 prompts, 1 brand
  • OpenAI web search sampled weekly
  • Share of voice and cited sources
  • 30-day history
Start free

Pro

Start here
$28/mo billed $340/yr · $34 monthly

Track your visibility over time, not just once.

  • 25 prompts shared across 2 projects
  • OpenAI web search · Google AI Overviews · Perplexity
  • 3 samples per prompt, per engine, every week — so a change is a trend, not noise
  • Up to 1,200 answers/mo · 90-day history
Start 7-day trial →

$2.00 test credit included

Growth

Most popular
$99/mo billed $1190/yr · $119 monthly

Weekly tracking on every engine.

  • 100 prompts shared across 5 projects
  • OpenAI web search · Google AI Overviews · Perplexity · Gemini · Claude
  • 3 samples per prompt, per engine, every week — so a change is a trend, not noise
  • Up to 7,800 answers/mo · 1-year history
Start 7-day trial →

$2.00 test credit included

Agency

$249/mo billed $2990/yr · $299 monthly

Client workspaces and white-label reports.

  • 100 prompts shared across 5 projects
  • OpenAI web search · Google AI Overviews · Perplexity · Gemini · Claude
  • 3 samples per prompt, per engine, every week — so a change is a trend, not noise
  • Up to 7,800 answers/mo · 2-year history
Start 7-day trial →

$2.00 test credit included

“Can’t I just ask ChatGPT myself?”

You can — and you should, for a gut check. But one manual chat tells you about as much as googling your own name once, logged in.

  • One chat is one sample. Models are non-deterministic; visibility is a rate — “named in 3 of 10 runs”.
  • Your own account flatters you. History, login and location leak in. We ask from clean, neutral sessions.
  • Buyers ask dozens of questions on several engines. Nobody repeats that by hand every week.
  • The value is the trend. GEO work takes weeks to land; you need logged history to see 12% become 34%.
The full answer →

51 free tools to fix what AI and search see

The fixes that move your score, each with evidence and copy-paste snippets. Free, no account.

Frequently asked questions

What counts as a prompt?

A prompt is one buying question you want to be recommended for — for example “best robot vacuum for pet hair”. Each cycle we ask it several times on every engine in your plan, from clean sessions, and record whether your brand was named, who else was, and which sources the answer cited.

Which AI engines do you track?

Claude, OpenAI web search, Perplexity, Gemini and Google AI Overviews. The free plan samples Claude; Pro adds OpenAI web search and Perplexity; Growth and Agency can use all five when configured.

How is share of voice calculated?

Share of voice is your brand mentions divided by all brand mentions (you plus the competitors you track) across a cycle's answers. “Mentioned in answers” is the share of answers that name you at all. Both are rates, not single samples, so they are comparable week to week.

Do the free tools stay free?

Yes. All 51 tools — AI readiness, llms.txt, DNS, SSL, email deliverability, security headers and the rest — stay free and need no account. Paid plans add what a one-shot check can't: tracking over time, competitors, alerts and reports.

Can I cancel, and is there a refund?

Cancel anytime from the billing portal; access runs to the end of the period. Eligible new accounts receive the trial shown for their plan, and if the first month isn't right, they can email us within 14 days for a refund.

Find out who AI recommends instead of you.

Three prompts, every week, free. Add competitors and engines when you’re ready.

Start tracking free