Utilitest Docs

Quick start

Copy the token’s contract address from your trading terminal, pump.fun or DexScreener.
Paste it into the box on the home page. The evaluation starts on paste.
Read the tier, the Why box and the claims. Every claim links to the source it was checked against.

Links work too: utilitest.fun/?ca=<address> opens a report directly, which makes reports easy to share.

Tip. Drag any evaluated token into your tier list and export it as an image.

Reading a report

Tier and score: The letter from S to F and the weighted score out of 100 behind it.
Headline and summary: What the project says it does, and what the evidence actually shows.
Why this tier: Two to four short points: what decided the grade, what held it back, and what would move it up.
Watch for: Promises that can’t be checked yet, such as milestone buybacks or launch dates. Re-run the evaluation later to see if they were kept.
Dimensions: Product, token role, usage and verifiability, each scored 0–100 with a one-line reason.
Claims: Up to five utility claims the project makes, each with a verdict and the evidence behind it.
On-chain & market: Mint and freeze authority, token age, supply burned since launch, liquidity, holder concentration, website, domain age and code activity. Green and red flags are listed underneath.
Sources: Every page, repository and data source the evaluation actually opened, numbered to match the citations.

Tiers

S88+Live product with real users, and using it requires the token, observably.

A75–87Working, used product with a specific token role.

B62–74Product works but the token role is thin, or the reverse.

C48–61Partial evidence.

D32–47Mostly promises, or a claim contradicted by evidence.

F0–31No utility claimed, or nothing that can be checked.

A contradicted claim caps the tier at D. The full rubric and weights are in the whitepaper.

An F on a pure meme isn’t an accusation. It means the token claims no utility, so there is nothing to credit.

Claim verdicts

Verdict	Meaning	Example
Verified	Evidence directly shows it.	“Mobile app” and the App Store listing is live with reviews.
Partial	Some support, with gaps.	Buybacks announced in a blog post; execution not shown.
Planned	The project presents it as future.	“Staking launches after migration.”
Unproven	No evidence either way.	“10,000 daily users” with no data anywhere.
Contradicted	Evidence conflicts with it.	“App is live” but the app link says coming soon.

Fresh tokens

Tokens under 14 days old are marked Fresh. They can’t have a usage history yet, so usage isn’t scored for them. Product counts 45%, token role 35% and verifiability 20%.

What helps a fresh token most is a product that works today, code and a domain that existed before the token, and token mechanics written down specifically: amounts, thresholds, triggers and how they’re enforced.

Full vs quick evaluations

Full evaluations have a language model read the whole evidence pack under the fixed rubric and check each claim. They take 20–60 seconds.

Quick evaluations are rules-based. They run instantly, detect documented token mechanics, and score what can be measured automatically, but they cannot judge whether a claim holds up. Quick evaluations run when your daily full evaluations are used up, or when the model is busy. The report always says which kind you are looking at.

Limits & access

Level	Hold ($UTEST)	Full / day	Deep scans / day	Watchlist
Free	—	1	—	—
Holder	100,000	10	2	5 tokens
Pro	1,000,000	40	10	15 tokens
Whale	5,000,000	150	40	40 tokens

Quick (rules-based) evaluations are unlimited for everyone. Every holder level also includes watchlist alerts and evidence report exports.

Holder features

Deep scan: Reads up to 18 website pages instead of 7, follows more of the sitemap, keeps more text from each page, and reads up to 10 of the project’s GitHub repositories. Always a full evaluation.
Watchlist alerts: Watched tokens are re-evaluated every 6 hours while Utilitest is open in your browser. You’re alerted when the tier or score moves by 5 or more, a claim’s verdict changes, mint/freeze authority is re-enabled, or the website goes down or comes back, with a browser notification if you allow it.
Export: Download any report as a PDF, Markdown or JSON (the complete raw evidence).

Verifying your holdings

Press Connect wallet in the $UTEST section and sign the message your wallet shows. Signing is free: it sends no transaction and moves no tokens. Utilitest reads your balance on-chain and re-checks it every few minutes, so your level always matches what you hold now. The session lasts 7 days.

Status. Holder levels switch on when $UTEST launches. Until then everyone is on the Free level.

Reports are cached: a full evaluation is reused for six hours (opening a cached report doesn’t use your daily allowance), a quick one for fifteen minutes. A cached report shows when it was produced.

For projects: improving your tier

Tiers can’t be bought or appealed. They change when the evidence changes. Utilitest reads what a careful outsider would read, so make that easy:

Link your website in the token metadata and in your DexScreener profile. Without a link, there is nothing to crawl.
Publish a tokenomics page at a clear address such as /tokenomics or /token. State exactly what the token does, with amounts, thresholds and triggers.
Keep a sitemap at /sitemap.xml that includes your docs, whitepaper and announcements.
Make the product reachable: a public app link, store listing, bot handle or demo.
Link your repository. Public code with recent commits is among the strongest product evidence there is.
Make mechanics verifiable on-chain: publish staking contract and buyback wallet addresses, and burn to reduce supply where you claim burns.
Mark future features as future. “Staking launches in Q1” is graded planned. “Staking is live” with no staking is graded contradicted, and that caps the tier at D.

API

The site is powered by a simple JSON endpoint. It is rate-limited like the site and may change before the public API is released.

GET /api/scan?mint=<contract address>
GET /api/scan?mint=<contract address>&deep=1   (holders: deep scan)
GET /api/access                              (your level and today’s usage)

The response has two parts: evidence, everything that was collected, and grade, the evaluation.

{
  "grade": {
    "tier": "B",
    "score": 67,
    "headline": "…",
    "why": ["…"],
    "watch": ["…"],
    "subjects": { "product": { "score": 85, "comment": "…" }, … },
    "claims": [{ "claim": "…", "verdict": "partial", "evidence": "…", "source": "https://…" }],
    "stage": "established",
    "mode": "ai"
  },
  "evidence": { "mint": "…", "stage": { "ageDays": 510, … }, "chain": { … }, "market": { … }, … }
}

GET /api/fresh returns recently launched tokens with a DexScreener profile.

FAQ

Is a high tier a buy signal?

No. Utilitest measures evidence of utility, not price, liquidity or returns.

Why did two scans of the same token differ?

One was probably a quick evaluation and the other a full one, or the project changed its site in between. Each report shows its method and when it was produced.

Why does an established project only get a B?

Usually because the product is real but the token is optional: the app works without it. A and S require the token to have a specific, enforced job.

Does Utilitest read X?

Not yet. Website, docs, code and on-chain data are read; X accounts are planned.