CopyGrade
Trust

We're paid by copiers, not by the wallets we score.

CopyGrade is an independent analytics firm. The flat subscription is how we stay honest — every other revenue model creates a reason to call someone copyable when they aren't. This page is the receipt.

How we make money

One revenue line. Everything else is a bright-line no.

Flat monthly subscription
$25/month or $20/month billed annually. That is the only revenue line. The number on a verdict has the same effect on our bank account whether it is 92 or 12.
No revenue share with any wallet
We do not collect a cut of any copier's PnL. A revenue-share model would pay us more every time we said 'copy' and nothing when we said 'avoid' — the worst possible incentive for an independent scorer.
No affiliate or referral fees from execution bots
We do not earn a commission when a copier sends a trade through a third-party bot. We name execution providers when relevant; none of them pay us to be named.
No paid placement on the leaderboard
A wallet cannot pay to rank higher, hide a farming flag, or buy a verdict revision. There is no order form for that and there will not be one.
What we don't do

The product is analysis. Everything that isn't analysis is somebody else's job.

We never execute a trade. The verdict is ours; the trade is yours.
We never custody funds, hold keys, or sign on anyone's behalf.
We do not give financial advice. A Copy Score is a research output.
We do not hold equity in any Polymarket wallet, fund, or operator we score.
We are not affiliated with Polymarket. No data partnership. No shared ownership. No referral arrangement.
How the score is held accountable

Every number on the verdict can be traced to a documented factor or a named trade.

Every score is decomposable
Read the methodology
The 0–100 number is the weighted blend of five named sub-scores. Click into any verdict and you can read every factor, every weight, and the on-chain trades behind any farming flag. If we can't show the work, we don't ship the number.
Every farming flag is forensic
See the farming taxonomy
Iceberg accumulation, decoy wallet clusters, stealth YES/NO merges, self-trade washing — each detector is defined with the exact threshold (trade count, window, edge ratio) that triggers it, and every flag cites the trades. No vibes. Only evidence.
The v1 model is heuristic — and we say so
See the track record
Thresholds (8 sub-$200 buys in 30 minutes, 8 near-price pairs in 10 minutes, 350 trades for <1.5% edge) are priors, tightened in a first pass against real activity. We snapshot every score and grade it against realized forward copier outcomes — the band-by-band results will publish on the methodology page's track record as each horizon matures (the first 30-day cohort matures late June 2026), whichever way they point, and the first weight retune waits on that window. False positives are possible and we document them.
The compact

If a verdict ever changes for a reason that isn't in the methodology, we've broken this promise.

Scoring rules change in public, with a dated note.
False positives are documented openly on the farming page.
No private revisions. No off-record verdict edits.
Disputes go to a real human: trust@copygrade.com.
Read the methodology before you pay us anything.
The free tier exists so you can audit the verdict on real wallets before you decide whether the score is worth $25/month.
How the score worksStart free