TapeXray vs Prediction Hunt: what actually differs in 2026
Prediction Hunt's arbitrage scanner is free and claims fee-adjusted profit calculation with a resolution-mismatch warning. TapeXray's board is also free and publishes the archive behind its numbers: 236,027 markets scanned, 10,869 pairs checked, and 2 gaps still positive after fees on 2026-09-23 - the same-priced differentiator is the data trail, not the price.
| Dimension | TapeXray (live) | They say | Source |
|---|---|---|---|
| Size of the archive scanned | 263,852 markets in the archive | Free scanner; the page states no market or archive count. | Source |
| Matched pairs scored against a published funnel | 8 matched pairs scored against the fee line | Free scanner claims fee-adjusted profit calculation but publishes no matched-pair count or funnel breakdown. | Source |
| Historical odds archive | 2026-09-26 archive last updated | No historical odds archive is mentioned on the page. | Source |
What does Prediction Hunt's scanner actually offer?
Prediction Hunt runs its arbitrage tool at prediction.com/arbitrage, read on 2026-09-23 (predictionhunt.com/arbitrage now redirects there with an HTTP 301). The scanner itself is free - no tier, no paywall on the page this desk read. It explicitly claims fee-adjusted profit calculation, meaning it says it accounts for venue fees before showing a spread, and it carries a specific warning about resolution-criteria mismatch: two markets can ask what sounds like the same question and still settle on different rules.
That warning matters, and TapeXray's own matcher exists to act on the same risk rather than just name it: a title match alone is not enough to call two markets the same bet. Prediction Hunt's page does not describe how many markets its scanner covers, how many candidate pairs it checks, or how many survive its own matching logic in a given day - none of that is published on the page this desk read, free or not.
What does TapeXray publish that Prediction Hunt's page does not?
TapeXray tracks 236,027 markets across both venues - 97,194 on Kalshi and 138,833 on Polymarket - and every stage of its own matching pipeline is published, not just the final spread. On 2026-09-23 the matcher checked 10,869 title-matched candidate pairs, rejected 7,218 as suspect, cleared 6 through the same-bet gate, and found 4 of those sitting inside the combined fee line with 2 still positive after fees. That funnel updates on the live board with every deploy, not once a day by hand.
| Stage | Count | 2026-09-23 |
|---|---|---|
| Candidate pairs matched by title | 10,869 | checked |
| Rejected as suspect | 7,218 | too weak, thin, or wide |
| Cleared the same-bet gate | 6 | scored |
| Inside the combined fee line | 4 | not worth trading |
| Positive after fees | 2 | real gaps today |
TapeXray also reads Polymarket's taker-fee rate per market from that market's own Gamma catalog row rather than assuming one number for every category, defaulting to the highest published rate when a schedule is missing or unreadable so an unknown fee never flatters a trade. Prediction Hunt's page states that it adjusts for fees but does not describe where each venue's rate comes from, so this is a statement about what TapeXray's own fee calculator shows its working for, not a claim about how Prediction Hunt computes its own number.
Is TapeXray actually cheaper than Prediction Hunt?
No, and this page will not pretend it is: Prediction Hunt's scanner is free, and so is TapeXray's board. Price is not the axis that separates the two products, and a comparison built on "we're cheaper" would be false on its face. The real difference this page can actually check is disclosure - how much of the archive, the funnel, and the fee sourcing behind a number is published for a reader to verify, rather than taken on the product's word.
TapeXray publishes its 236,027-market archive, its 10,869-pair funnel, and its per-market fee sourcing on the live site, all rebuilt every six hours. Prediction Hunt's own arbitrage page states its two headline claims - fee adjustment and a resolution-mismatch warning - without publishing the underlying counts this desk could check them against. See ArbBets for the other product verified on the same date, and the markets archive for the full catalog behind TapeXray's own numbers.