What’s the catch? Nothing we could corroborate against this app.
Analyzed automatically from 3 independent signals that had to agree — not a hands-on test, and not human-reviewed.
This analysis describes version 1.5.6 of the app, analyzed 2026-10-04 . The app may have changed since.
The store currently lists version 1.5.6, updated 2026-08-13.
What the analysis found
Context 1
Measured and published because a reader would want to know — deliberately not a charge. The measurement alone does not show a harm.
One source only — shown, not confirmed 1
A single kind of evidence. We publish these because withholding them would be the bigger distortion, and we do not treat any of them as a finding we stand behind.
What reviewers said
Verbatim excerpts from public store reviews, shown as corroborating user reports. They are not Appworthy audit evidence, and a quotation is not proof of the finding it sits under.
Aggressive ad load
I love the app, — In my opinion, the bots memory isn’t as bad as people say it is. Like most bots I’ve talked to remember small details even after multiple messages and that’s something pretty impressive to me. This app is really similar…
worst ai app ever, id give it a zero star if i could — a lot of ads and waiting 14 minutes just to chat is absurd. also the daily message limit. Not to mention that the language of the messages when i’m chatting randomly changes to…
Great but too many adds — There are so many ads I love the app but I can barely get 4 messages in before it shows a ad
What this app sells
From the store listing. Billing terms are read from each purchase’s name — the stores don’t publish them — so treat them as a strong hint, not a disclosure.
How the automated analysis works
This assessment is produced automatically. Three independent signals are read — user reviews, the store listing and the store privacy disclosure — and each candidate finding is filed under the sources that support it.
No charge stands on one kind of evidence — with one published exception. A finding is either a measured fact that anyone can reproduce (an ad SDK is in the package; every subscription bills weekly; the privacy label omits a tracker we can see), or it is an interpretation — and an interpretation needs independent sources to agree before it reaches the catch list. The exception is a harm only users can see: a charge may publish from reviews alone, as “Reported by N users”, once enough reviewers raise it to clear that charge’s own published bar. That agreement rule is what replaced the human reviewer, not a model’s confidence.
There is no overall score, on purpose. A 0–99 number was published here until 2026-09-15. It was withdrawn because it discriminated too weakly to be the instrument — 44.7% of every pair of apps tied — and because a single number is a verdict where a count is an observation. What publishes instead is a row of measurements, each of which says what was counted and, where we could not look, says that instead.
Findings under One source only are real reports we do not stand behind as findings. They are shown because withholding them entirely would hide a genuine complaint, and labelled because publishing them as charges would be an accusation off a single source.
Several measurements above say “Not assessed” because no artefact we hold can answer them. A human review is what resolves those — and it can remove a favourable band as easily as add one.
Hands-on inspections are not yet open.