About DXT Tools

What this site is

DXT Tools is an independent tracker of paid activity on DexScreener: every time a token pays for a Dex Paid profile (Enhanced Token Info), buys a boost, runs an ad campaign or gets a community takeover, we record it within seconds — with the market cap at that moment. We then follow what happens: the highest market cap reached after the payment, whether it is still up, and whether it flopped.

The name is DXT Tools — also written DXTTools or dxttools — and the site lives at dxttools.trade. It is not related to DEXTools (dextools.io), to the DirectX "DXT" texture tools, or to Anthropic's "DXT" desktop-extension format.

The live app is at dxttools; the DexScreener guide explains how the platform's paid services work and what the data says about them.

How the data is collected

• The public DexScreener API is polled continuously (token profiles, boosts, ads, pair data). Nothing is scraped from the website.
• Each event is stored with a snapshot of price, liquidity and market cap taken at discovery ("Paid @").
• "ATH after paid" combines our own minute-by-minute samples with historical candles from GeckoTerminal, and is marked final 7 days after the payment.
• "Flop" means the token trades at ≤30% of the cap it paid at, judged 30+ minutes after the payment. "Seen before" flags tokens sharing links, descriptions or images with other paid tokens.
• Events are kept for 180 days; token summaries stay so a contract can always be checked.

Independence

DXT Tools is an independent resource. It is not affiliated with, endorsed by or connected to DexScreener. "DexScreener" is a trademark of its owner; it is used here only to describe the platform. Nothing on this site is financial advice. Trade-button links carry our referral codes; that is how the tracker stays free. We do not accept payment to change data, rankings or verdicts. Sponsored placements are labelled.

Privacy and measurement

We count visits ourselves. Our own measurement uses no cookies and stores nothing on your device.

What the statistics database holds:

• Page views: the page (for a token page, the token's address and chain), the time, whether the row is a request, a "page was opened" signal or a "first input" signal, and a visitor hash — the first 16 characters of a SHA-256 over a random daily salt, your IP address (for IPv6, its /64 prefix) and your browser's user-agent. The salt is replaced at 00:00 UTC, so the same visitor has a different hash every day; the previous day's salt is deleted one day later. Until early October 2026 the salt was kept inside the database, so a database backup made before that change still contains the salt of its day; the server's routine backups are kept for about two weeks.
• Whether the user-agent looks like a bot (yes or no), the hostname of the site you came from (not the full address), and a source tag when the link you followed carried one (?src= or utm_source).
• Searches typed into the site: the text, with the same daily hash.
• Trade-button clicks: the platform, the token, the kind of page, the daily hash, a coarse browser class such as "Chrome on Windows, desktop", and whether the click came from a real input event.

When a page has loaded, your browser sends a small request saying that the page was really opened in a browser; in the app, another one is sent when you switch to a different tab of the app. The body carries the page path and, on the first such request of a page load only, the referring hostname and the source tag. One more request is sent the first time you touch, click, press a key or turn the mouse wheel on a page (once more if you return to the page with the browser's back button): it carries the page path and the fact that an input happened — not which key, not what you clicked and not where. Nothing else is in these requests.

The statistics tables contain no IP address and no user-agent string. Page-view and search rows are deleted after 180 days, click rows after 365 days.

Separately, the web server (nginx) in front of the app writes a standard access log for security and debugging: IP address, time, request, referrer and user-agent. It is rotated daily and 14 rotated files are kept, so an entry is deleted 14 to 15 days after it was written. For rate limiting and connection limits, the app also keeps IP addresses in memory: for a few minutes after a request, and for as long as an open tab's live-update connection lasts. An address with repeated failed sign-ins to the admin panel is kept in memory for up to half an hour and is written to the application log.

The app keeps your own settings in your browser's local storage: the trade-button order, the view mode, alert settings, your watchlist, the time-zone switch, the Hot strip's settings, and whether you closed the intro strip or a site notice. These are preferences, not identifiers: they are not read for measurement and are not sent with the measurement requests.

We also load the analytics script of Umami Cloud, a third-party service, on our pages. Your browser sends it the page address and title, the referrer, your screen size and your language, and — as with any web request — your IP address and user-agent. According to Umami's documentation it sets no cookies and does not store IP addresses; it derives the country, browser, operating system and device type from them. Content blockers commonly block this script, which does not affect the site.

Pages load fonts from Google Fonts and token images from third-party image hosts, and when our own price budget is used up your browser may ask DexScreener's public API for prices directly. Those hosts receive your IP address and user-agent, as any web server does.

Trade buttons are referral links. A click is recorded as described above, and the platform you land on applies its own privacy policy.

Contact

Questions, corrections, partnerships: Telegram or @dxttools.