Guide
What a Vetall trust score means, band by band
Vetall puts a number from 0 to 100 next to a Reddit username. This page explains exactly what that number is, what each band means, and, just as important, what the score cannot know. A trust product owes you total honesty about its own limits.
Where the score comes from
When an account is checked, its recent public history, typically a couple of hundred comments and posts, is read and reduced to measurements: how old the account is, how its karma balances between posts and comments, how regularly it posts, how much of its text repeats, how much has been removed by moderators, and how widely it roams across subreddits. An AI then weighs those patterns together into a score, a band, the top signals that drove the verdict, and one line of advice. No single signal ever decides the outcome.
The text itself is discarded after scoring; only the measurements and the verdict are kept. And the score describes observable public behaviour, never who someone is: the fixed disclosure under every verdict says exactly that.
The four bands
| Band | Score | What it says, and what to do |
|---|---|---|
| Trusted | 80 to 100 | A long, consistent, good-faith record. Engage normally; ordinary caution still applies to money. |
| Neutral | 55 to 79 | Unremarkable or mixed signals. Most accounts live here. Nothing alarming, nothing vouching. |
| Caution | 30 to 54 | Concerning patterns worth taking seriously. Verify claims independently before trusting. |
| High Risk | 0 to 29 | Strong scam, bot or shill signatures. Do not act on financial or marketplace claims from this account. |
Confidence, and the courage to say "not enough"
Every verdict carries a confidence level. Blatant patterns produce very certain verdicts; sparse ones do not. And when an account is simply too thin to judge, a two day old account with three comments, the verdict says exactly that instead of dressing a guess up as a number. A score that cannot admit uncertainty is not a score, it is a horoscope.
What the score cannot know
- Private behaviour. Direct messages, deleted history and anything behind a login are invisible to it, by design.
- Intent. It reads patterns, not minds. A "High Risk" verdict means the account behaves like accounts that scam, not that a court would agree.
- Stolen accounts. A hijacked aged account inherits its victim's clean record until its behaviour changes. Patterns catch the change, not the theft.
That is why every verdict ends with the same sentence: an AI-generated assessment of public activity patterns, not a statement of fact. Treat the score the way you would treat a knowledgeable friend's read on a stranger: genuinely useful, worth weighing, and never a substitute for a payment method you can reverse.
Who sees what
- Anyone with the extension sees the band of every account somebody has already checked, free, forever. One person pays for a check; everyone benefits.
- Your own checks show the full picture: the exact score, the signals, and the advice. Every account starts with 7 free checks, and anything you check stays unlocked for you permanently.
- Deep dives, on the Analyst plan, write the full multi-paragraph read of an account's history, including what to verify before acting on its claims.
See a score on a real thread.
Install free, open any Reddit thread, and the badges appear. 7 full checks included to run on any account you meet.
Add Vetall to Chrome, freeNo card required. Cached badges are free on every plan, forever.