Neuronto PaymentsAccept payments

Status

98.05% availability over the last 30 days (193 probes, 0 failed, median path 224 ms). That figure is the lower bound of the 95% confidence interval, not the raw rate (100.00%), so it understates rather than flatters.

Networks right now

NetworkCAIP-2KindState
Base Sepoliaeip155:84532testnetpaused
Baseeip155:8453mainnetaccepting payments

A network is withdrawn from /supported before it stops settling, so a

client that reads /supported never sends a payment that cannot land.

Settlements

WindowSettled
Last 24 hours0
Last 7 days0
Last 30 days0
All time0

Pending right now: 0. Failed in 30 days: 21. Catalogued resources:

0.

Screening

Payments to or from addresses on the OFAC SDN crypto addresses are refused: 124 addresses, last updated 2026-09-17T07:10:06Z (0.6 hours ago).

The screen covers addresses against one public list. It is not a name screen, not a jurisdiction

check and not a compliance programme, and a refusal is returned as sanctioned_address before any

chain call, so a refused payment costs nobody gas. If the list is missing or out of date, payments

continue and the age above says so: a file that failed to download is a poor reason to stop

settling every payment, and hiding that state would be worse than either.

How this is measured

A probe runs every ten minutes beside the service and records one row per attempt: the whole

public path, not an internal health check. Availability is computed from those counters, so the

denominator is real. Under 5 probes the count is published and the rate is withheld.

Above it, the published figure is the Wilson lower bound.

The honest gap: an outage that also stops the prober records nothing rather than recording a

failure, so this measurement can only understate an outage of the whole host, never overstate

availability within a running host. Anyone who wants an independent number should probe

/health themselves; it is free, unauthenticated and rate limited

generously enough for a monitor.

What this is not

This is not a service level agreement, and there is no uptime promise. The figures are what was

observed, published whether they flatter us or not, and the raw counts are in

/status.json for anyone who wants to compute their own.