What They Advertise vs. What Their Own Pages Show
Every row below is two verbatim quotes taken from the same reseller's own published pages, each with its own capture date and a link back to where we found it. We don't infer motive and we don't calculate anything from either side — we just put them next to each other. See our methodology for how we handle quotes generally.
| Seller | What it advertises | What its own pages show | Captured |
|---|---|---|---|
| TokenMart (thetokenmart.ai / .com) | “up to 65% below retail”
meta description — source |
“SAVE UP TO 15%”
homepage discount ticker banner — source |
2026-09-23 |
| CheapAI (cheapai.io) | “up to 87% cheaper than retail”
homepage headline — source |
“determined by the Model Provider's rate plus the Operator's minimal markup”
terms of service, Definitions section — source |
2026-09-23 |
| CheapCredits.ai | “AI API Tokens at 40% Off”
page title — source |
“40–60% below list price”
meta description — source |
2026-09-23 |
What each pair is actually saying
- TokenMart (thetokenmart.ai / .com): Both lines are on the same homepage, captured the same day: the page's <meta> description still says up to 65% off, while the scrolling ticker banner a few lines into the same HTML says 15%.
- CheapAI (cheapai.io): The homepage headline advertises a discount off the official price; the terms define the price per request as the model provider's own rate plus the operator's markup.
- CheapCredits.ai: The site is a single one-page listing with no pricing table, so these two numbers — 40% in the browser tab's title, 40–60% in the same page's meta description — are the only discount figures on its one public page, and they disagree with each other.
Everything on this page is a verbatim quote from a provider's own published page, with a capture date and a link to the source. We keep an archived copy of every page quoted here; if a provider updates a page and the contradiction goes away, we'll record that change on our terms changelog rather than just deleting the row with no note.