Transparency

Service status

This page states what works, what does not, and what we do not measure yet. It lists one component per payout method and per blockchain network, because “the service is up” means nothing when it is your rail that is not moving.

Page rendered on 5 September 2026 at 13:1944 components tracked

No automated measurement is in place yet

Not measured

No probe writes into this page yet. We could have shown a green bar and a “99.9%”: it would be invisible to you, indefensible if you asked, and untrue. Until the measurement exists, the state shown is “not measured”, and nothing else.

How this page will be fed

The mechanism is described before being connected, so you can judge its worth when it is. A status page fed by the service it watches goes down with it: the probe will therefore be external.

Why no availability percentage is displayed

An uptime figure only means something backed by a method: which probe, at what frequency, from where, counting what as an outage. Without that, “99.95%” is set dressing. We will not publish one before we have the measurement, and we will publish the method alongside the figure.

The trade-off is deliberate: this page is less flattering today than a wall of green. It has the advantage of being true, and of staying true the day an incident happens.

Probe: Not connected
Monitoring provider: none selected to date
Health endpoint: /api/health
Target sampling period: 60 s
Published window: 90 days
Last sample: none
Measured components: 0 / 44

Components

For each component, the “what is checked” column describes the exact control. That is the contract: the day the dot turns green, it will mean what that column says, and nothing more.

Platform

The site, the API and account access. This group rarely fails alone: when it does, everything else follows.

Platform
Component	What is checked	90-day availability	State
Public website	GET on the home page in both locales, HTTP 200 and full render of the quote widget.	measurement being put in place	Not measured
API	POST /api/quote with a known parameter set, checking HTTP 200 and that the fee breakdown balances.	measurement being put in place	Not measured
Accounts and sign-in	Sign-in with a dedicated monitoring account, then reading the dashboard.	measurement being put in place	Not measured
Identity verification	Availability of the verification provider and median decision time on automated cases.	measurement being put in place	Not measured

Rates and quoting

A silent rate source is not a cosmetic degradation: the engine refuses to quote rather than serve a stale price, and the button goes inactive. That is deliberate.

Rates and quoting
Component	What is checked	90-day availability	State
Rate sourcevia CoinGecko	Age of the latest mid-market tick per pair. Past the maximum tolerated age the engine refuses to quote: that is an incident, not a slowdown.	measurement being put in place	Not measured
Quoting engine	End-to-end quote on three reference corridors, checking the gross minus fees equals net invariant.	measurement being put in place	Not measured

Network watchers

A deposit invisible to the customer is nearly always a gap between our block height and the network’s. Every network accepted for deposit is watched separately.

Network watchers
Component	What is checked	90-day availability	State
Bitcoin watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 2 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured
Ethereum (ERC-20) watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 12 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured
Tron (TRC-20) watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 20 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured
Solana (SPL) watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 32 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured
BNB Smart Chain (BEP-20) watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 15 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured
Polygon PoS watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 128 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured
Arbitrum One watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 20 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured
Base watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 20 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured
OP Mainnet watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 20 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured
Litecoin watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 6 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured
XRP Ledger watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 1 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured
Stellar watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 1 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured
TON watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 1 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured
Avalanche C-Chain watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 20 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured
Dogecoin watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 20 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured
Bitcoin Cash watcher	Gap between our block height and the public explorer, plus detection delay on a canary deposit. Alert threshold: 6 confirmations not reached within twice the theoretical time.	measurement being put in place	Not measured

Payouts by method

This is the level of detail that matters to you: “the service is up” means nothing if it is your rail that is not moving.

Payouts by method
Component	What is checked	90-day availability	State
Virement SEPA Instantanevia bvnk, modulr, banking-circle	24h payout success rate, median delay compared with the announced p50 (1 min) and rejection rate by reason.	measurement being put in place	Not measured
Virement SEPAvia bvnk, modulr, banking-circle	24h payout success rate, median delay compared with the announced p50 (240 min) and rejection rate by reason.	measurement being put in place	Not measured
Faster Paymentsvia modulr, clearbank	24h payout success rate, median delay compared with the announced p50 (2 min) and rejection rate by reason.	measurement being put in place	Not measured
ACHvia column, increase	24h payout success rate, median delay compared with the announced p50 (1440 min) and rejection rate by reason.	measurement being put in place	Not measured
Domestic Wirevia column, increase	24h payout success rate, median delay compared with the announced p50 (180 min) and rejection rate by reason.	measurement being put in place	Not measured
Interac e-Transfervia nuvei, peoples-trust	24h payout success rate, median delay compared with the announced p50 (5 min) and rejection rate by reason.	measurement being put in place	Not measured
PayPalvia paypal-payouts	24h payout success rate, median delay compared with the announced p50 (10 min) and rejection rate by reason.	measurement being put in place	Not measured
Wisevia wise-platform	24h payout success rate, median delay compared with the announced p50 (30 min) and rejection rate by reason.	measurement being put in place	Not measured
Revolutvia bvnk, modulr	24h payout success rate, median delay compared with the announced p50 (2 min) and rejection rate by reason.	measurement being put in place	Not measured
PIXvia dlocal, transfero, ebanx	24h payout success rate, median delay compared with the announced p50 (1 min) and rejection rate by reason.	measurement being put in place	Not measured
SPEIvia dlocal, bitso-business	24h payout success rate, median delay compared with the announced p50 (3 min) and rejection rate by reason.	measurement being put in place	Not measured
Nequivia dlocal, koywe	24h payout success rate, median delay compared with the announced p50 (5 min) and rejection rate by reason.	measurement being put in place	Not measured
Transferencia CBU / CVUvia dlocal, lemon-cash	24h payout success rate, median delay compared with the announced p50 (5 min) and rejection rate by reason.	measurement being put in place	Not measured
M-Pesavia onafriq, thunes, kotanipay	24h payout success rate, median delay compared with the announced p50 (2 min) and rejection rate by reason.	measurement being put in place	Not measured
Virement bancaire NIPvia flutterwave, paystack, onafriq	24h payout success rate, median delay compared with the announced p50 (3 min) and rejection rate by reason.	measurement being put in place	Not measured
MTN Mobile Moneyvia onafriq, flutterwave, thunes	24h payout success rate, median delay compared with the announced p50 (3 min) and rejection rate by reason.	measurement being put in place	Not measured
Wavevia onafriq, thunes	24h payout success rate, median delay compared with the announced p50 (3 min) and rejection rate by reason.	measurement being put in place	Not measured
Orange Moneyvia onafriq, thunes	24h payout success rate, median delay compared with the announced p50 (5 min) and rejection rate by reason.	measurement being put in place	Not measured
GCashvia thunes, dlocal	24h payout success rate, median delay compared with the announced p50 (3 min) and rejection rate by reason.	measurement being put in place	Not measured
Virement FASTvia thunes	24h payout success rate, median delay compared with the announced p50 (5 min) and rejection rate by reason.	measurement being put in place	Not measured
Virement bancaire AEDvia nium, thunes	24h payout success rate, median delay compared with the announced p50 (240 min) and rejection rate by reason.	measurement being put in place	Not measured
Virement international SWIFTvia banking-circle, nium	24h payout success rate, median delay compared with the announced p50 (2880 min) and rejection rate by reason.	measurement being put in place	Not measured

Last 90 days

One cell per day. A grey cell means “not measured”, not “no outage”: that nuance is the whole difference between a status page and a decoration.

2026-06-08Grey: day not measured2026-09-05

No incident published

No incident has been published since this page opened on 2 September 2026. That sentence says exactly what it says: no incident published. It does not certify that none happened — which is precisely what the probe will let us assert.

Incident report template

This is the level of detail we commit to for a real incident: quantified scope, named technical cause, and what was changed so it does not happen again. The template is published before the first incident, precisely to avoid writing it in the heat of the moment, with the temptation to downplay.

EXAMPLE — report formatting. This incident never happened.

SEV2

SEPA Instant payouts queued for 1 h 48

2026-09-02T09:14:00Z → 2026-09-02T11:02:00Z

What it changed for you: Payouts to euro-area IBANs stayed queued instead of leaving in under 5 minutes. No order was lost and no rate was recalculated: the locked rate remains the one set when the order was created, incidents included. Deposits, quoting and the other rails were unaffected.
Cause: Our partner institution rotated a client certificate without prior notice. Our transfer calls were rejected at the TLS layer; the payout queue correctly held the operations instead of failing them, which prevented any double send.
What was fixed: Certificate updated at 10:51, queue drained in 11 minutes. Three changes since: expiry monitoring on every provider certificate 30 days ahead, an alert as soon as a payout queue exceeds twice the rail p95, and automatic failover to the rail’s second provider after 15 minutes of consecutive failures.

Timeline

2026-09-02T09:31:00Z · Investigating
We see SEPA Instant payouts failing to leave. Deposits and quoting are working normally. Other rails are not affected.
2026-09-02T10:05:00Z · Cause identified
Cause identified: TLS rejection on the partner institution side after a certificate rotation. Operations are held in the queue, none lost or sent twice.
2026-09-02T10:51:00Z · Monitoring
Certificate updated, the queue is draining. We are watching for a return to normal delay before closing.
2026-09-02T11:02:00Z · Resolved
Queue empty, delay back under 5 minutes. Incident closed. Full report published on this page.

Severity policy

These delays are internal commitments, not measurements. They are deliberately achievable: an announced delay that is missed costs more than a long one that is met.

Level	Definition	First update	Cadence	Report
Majorsev1	Customer money is stuck, or an order cannot progress: deposits undetected, payouts halted on several rails, quoting impossible.	15 min	30 min	5 days
Partialsev2	One rail, network or function is down or clearly slower than its p95, the rest works.	30 min	60 min	10 days
Minorsev3	Visible degradation with no consequence on funds: slow rendering, delayed notification, a secondary page erroring.	120 min	240 min	not systematic

An incident caused by a provider is published here like any other. From where you stand, the difference between “it is us” and “it is our partner” is zero: your money is late either way.
An incident is not closed until its cause is written down. “Restored” with no explanation is a polite way of saying nothing.
Incident reports stay readable for 24 months. We do not delete a past incident.

External dependencies

Most outages on a service like this one come from outside: a payment institution, a congested blockchain, a silent price source. Better said upfront.

Rails with a single provider

A rail served by one provider has no fallback: if that provider goes down, the rail goes down. Those rails stay in beta or get a second provider before opening.

PayPal — paypal-payouts
Wise — wise-platform
Virement FAST — thunes

What does not depend on us

Congestion on a blockchain lengthens the confirmation time of your deposit. We see it, we display it, we cannot speed it up.
A receiving bank can hold an incoming transfer for its own checks. The rail says “settled”, your account is not credited yet: that last metre is out of our hands.
An unavailable price source disables quoting. That is a choice: quoting on a stale price would have us promise an amount we could not hold.

Scheduled maintenance

Predictable work is announced here in advance, with the components affected and the estimated duration.

No scheduled maintenance is announced at this time.

Getting notified

There is no notification feed or email subscription yet: building an alert channel before having a probe would mean promising alerts that would never be sent. In the meantime this page is the source, and the changelog records substantive changes.

[email protected]Changelog