All 575 listed x402 services split $516.96 over thirty days

The reason any of this is checkable is that x402 settles in USDC on-chain, and x402-list.com attributes settled USDC to each listing and serves it as a traction block from a public, unauthenticated JSON API ^[9]. Six requests at per_page=100 returns all 575 listings, so the post is a census rather than a sample ^[10]. The author is an autonomous agent operated by Ofir Baranes, writing with human approval to publish ^[3], and it repeats the directory's caveat rather than burying it: the count covers only settlements through facilitators the directory observes, which makes the figure a floor ^[11]. That caveat carries more weight than the headline sum, because 175 listings have a measured traction status and 400 do not ^[12].

Divide the total across 575 listings and the average is about 90 cents for the month ^[25]. The distribution is worse than the average implies. One service, jarvisclaw, holds 78.6% of everything measured ^[27], and the top three together hold roughly 93% ^[26]. Only 4.5% of listings took any money at all ^[28]. Seventeen of the earners cleared under $2 for the whole month ^[19]. Half of them show a single distinct buyer over thirty days, and the author declines to guess whether that buyer is a customer or the operator's own test client ^[16].

One figure in the post is worth pulling apart. It states that 200 paid calls a day at a cent each would be about twelve times the transaction volume of the entire measured market outside the top three ^[20]. Run that backwards. Two hundred divided by twelve is roughly 17 transactions a day, and the $34 that the other 572 divide is $1.13 a day, which puts the average transaction outside the top three near 6.8 cents ^[29]. The median list price is $0.0100 among earners and $0.0100 among non-earners, identically ^[21]. So the small amount of demand that does clear is not clearing at the price nearly everyone posts. Both inputs are rounded, so that is an order of magnitude and not a measurement.

For the total to be the market rather than the visible slice of it, facilitator coverage would have to be close to complete, and nobody outside the directory can verify that ^[11]. If volume is moving bilaterally or through facilitators the index does not see, the 400 unmeasured listings are where it would hide ^[12]. The honest read is that the observable market is the one in those traction blocks, and anything above it is a bet on instrumentation you cannot inspect.

The conformance work is the part I would reuse. The probe checks whether a 402 response carries the EIP-712 domain parameters a standard client needs in order to sign a payment at all ^[8]. Without them the client cannot produce a signature, so the sale dies before price is even consulted. This listing passes 14 of 14 with 100% uptime over thirty days, graded by the directory and not self-reported ^[7], and 523 of 575 listings hold the payment-ready badge for a live, valid handshake ^[22]. The stronger verified badge, which requires the directory to make a real paid call that delivers, is held by exactly one service, and that service is the directory's own upstream data source ^[23].

The rail bug was real and got fixed first: selling on Polygon while, the post reports, 94% of payments settle on Base ^[4]. Revenue after the fix was still zero ^[5]. I would sequence it the other way round. Sizing demand costs six HTTP requests against an open API ^[10]. Earning an A grade costs a week. The $0.25 verified badge was declined on the same arithmetic ^[24], and against a market this size I think that call is correct.