Skip to content
How it works · $5 free credit · no subscription

Submit prompts. See who AI engines cite. Pay only for what you check.

AuditAE runs every (prompt × engine) pair through ChatGPT, Perplexity, Gemini and Google AI Overviews in parallel — and tells you whether your brand was named, who got the link credit instead, and how the answer framed you.

No card required · $5 credit lands instantly
AEBOT · sharpbook.io$9.64 left
Send April’s full report to aaron@auditae.app now.
Reading audits GA4 sessions Search Console rank history technical audit saving report sending email
Done. April 2026 report saved and sent to aaron@auditae.app. Highlights:
  • AI citation rate +14% month over month
  • Two tracked keywords moved into the top 10
  • INP regressed on mobile — 312 ms (was 198 ms)
April 2026 · sharpbook.ioSaved · Sent
Want me to make this recurring on the 1st of every month?
$0.27 used · $9.64 left · replies generated by Claude using your audit + billing data
01The flow

From prompts to a report in three steps.

No onboarding call. No CSV upload. The first audit runs without an account — and the second one tells AEBOT to email the result to your client.

01

Add a site, drop in prompts

Pick the brand and domain you want measured. Paste 5–10 questions your buyers actually ask — full sentences, not keywords. Save the competitors you want tracked alongside you so share of voice is measured on every run.

02

Run across four engines

Every prompt × engine pair runs in parallel. Five prompts × four engines = 20 checks, ~3 minutes, $1.00. No queues, no batching, no waiting on a weekly refresh.

03

Read the gap

Per-prompt grid (cited / position / sentiment), share-of-voice panel, sources panel, and a competitor split that shows where the link credit actually went when it didn’t go to you.

02The engines

Four engines. All checked. No starter pack.

Each engine sees a different web. AuditAE hits all four — none are upsells, none gated behind a higher tier.

GPT
ChatGPT
Powered by OpenAI Responses API + web_search

Prose-heavy answers. Sometimes lists 3–5 sources, sometimes none. The hardest engine to move because brand recall from training data weighs heavily — unlinked mentions in places ChatGPT’s training set ingested matter as much as backlinks.

More on ChatGPT →
PPX
Perplexity
Powered by Sonar API

Most source-rich of the four. Usually 5–10 cited URLs per answer. Fastest engine to move with a content rewrite — fresh, on-topic pages get picked up within days. If you’re going to see early wins, they show up here first.

More on Perplexity →
GEM
Google Gemini
Powered by @google/genai with googleSearch grounding

Tracks Google Search rankings closely. Often quotes Wikipedia and Reddit. If you rank well in Google for the underlying query, you’re already mostly there — strong traditional SEO is the highest-leverage AEO investment for Gemini specifically.

More on Gemini →
AIO
Google AI Overviews
Powered by live SerpAPI (Google has no official AIO API)

The SERP-embedded answer above the blue links. Sensitive to E-E-A-T signals — author bylines, structured data, site authority, freshness. The most schema-responsive of the four engines.

AI Overviews tracker →
03Inside a cell

What every check tells you.

One cell = one prompt running on one engine. Every cell stores all of this, so you can roll the data up however you want — citation rate, share of voice, sentiment trend, competitor split.

Brand-mention detection is deterministic regex — case-insensitive, word-boundary-aware, prefix-variant-handling. Competitor extraction and sentiment use Claude Haiku. Source-domain matching is hostname-based. The full pipeline is documented in our methodology.

Stored per prompt × engine cell
cited

Did the engine actually link to your domain? Boolean.

position

Where your link sat in the engine’s source list. Lower is better.

sources

Every URL the engine cited for that answer.

competitors

Brands named in the answer prose, extracted by Claude Haiku.

sentiment

How the engine framed you — positive, neutral, or negative.

excerpt

The relevant chunk of the engine’s response, markdown-stripped.

cost_usd

Exact provider cost for the cell. Audit cost = sum of cells.

AI visibility report · yoast.com5 prompts · 20 checks
95%
Mentioned in answer
URL cited as source
EngineURL cited
ChatGPT3 / 5
Perplexity2 / 5
Gemini2 / 5
AI Overviews2 / 5
10 prompts mentioned the brand but sourced a third party — the engines know you; the link credit isn’t reaching your site.

What that looks like as a report.

A real audit: 5 prompts × 4 engines against a well-known SEO brand. Mentioned in nearly every answer — but the link credit only reached its own domain 45% of the time. That gap — mentioned vs. cited — is the whole game.

  • Mentioned vs. URL-cited, split per engine
  • Share of voice across every cited domain
  • Per-prompt breakdown with sources and sentiment
04Plus your data

Bring your Analytics + Search Console. AEBOT reads the rest.

Connect GA4 and Search Console once per site. Read-only by default — AEBOT cannot modify your Google properties or read your inbox. Per-site OAuth keeps different clients’ Google accounts isolated.

Google Analytics 4

Data API · runReport · analytics.readonly

Any dimension, any metric the Data API exposes — events, engagement, conversions, revenue, device and channel splits, custom dimensions. Not just sessions and top landing pages.

query_ga4(dimensions=["eventName"], metrics=["eventCount","totalUsers"])query_ga4(dimensions=["sessionDefaultChannelGroup"], metrics=["sessions","totalRevenue"])

Google Search Console

searchAnalytics.query · webmasters.readonly

Full searchAnalytics surface — query, page, country, device, searchAppearance. Regex filters, 90 days deep. The default window respects GSC’s 2-day data lag automatically.

query_gsc(dimensions=["page"], orderBy="clicks", rowLimit=10)query_gsc(dimensions=["query","page"], filterGroups=[regex impressions > clicks])
05Three ways in

Use the dashboard. Or skip it.

Same primitives, same billing pipeline, three front doors. Pick the one that fits your stack — or use all three.

Dashboard

Sites, audits, trackers, AEBOT chat, exports. Built for marketers and agencies who don’t want to write code. Every primitive surfaces here.

Try the dashboard →

REST API

REST endpoints for run_audit, check_citation, extract_competitors, run_technical_audit. Same billing pipeline as the dashboard. Keys in Settings.

Get an API key →

MCP server

Wire AuditAE into Claude Desktop, Cursor, or any MCP client. Same primitives, exposed as native MCP tools. Email support for the config.

Email for setup →
Ready when you are

Run your first audit. See where you land.

One prompt. Four engines. About 30 seconds. No card required — your $5 starter credit lands the second you sign up.

no card required · $5 credit lands instantly