How Long a Crypto Payment Actually Takes, Measured Across 1.48 Million Orders

A shopper picks the crypto option, the invoice opens, and then you wait. How long?

Search for an answer and you get block times. Bitcoin is ten minutes, Solana is under a second, Ethereum is twelve seconds. None of that is the number a merchant feels, because a payment is not a block. It is a person choosing a coin, opening a wallet, approving a transfer, and only then a chain doing its part.

So we measured the thing itself. Over the twelve months to the end of August 2026 we timed 1.48 million paid orders across fourteen networks, and split the wait into the part your shopper controls and the part the network does.

Some of what came out was predictable. Some of it was not.

Want to see this on your own checkout? Open a CoinGate account.

Two clocks, and only one of them belongs to the blockchain

An order here runs on two timers, which is what makes this measurable at all.

In the new status, the shopper has two hours to pick which currency and network they want to pay with. Once they pick, the order moves to pending and a second, much shorter clock starts. Twenty minutes to make the transfer.

That split lets us separate the human part of the wait from the machine part. The human part turns out to be the dull one.

Median time to choose a currency and network: between 29 and 61 seconds, on every network we measured. Arbitrum shoppers were quickest at 29 seconds, Bitcoin Cash shoppers slowest at 61. Half a minute of spread across the entire book. Whatever is making your crypto payments slow, it is not the selection screen.

Everything worth looking at happens after that.

The payment clock, network by network

From the moment the shopper picks a network to the moment the order is paid.

One note on the last row before anything else. Bitcoin Cash was payable for ten of the twelve months and the measurement is real, however it is no longer selectable at our checkout. It stays in the table as the honest worst case, not as an option you need to weigh.

Read the median column top to bottom and the shape is hard to miss. Lightning is done in 46 seconds. Bitcoin takes 14 minutes and 16 seconds. Same asset underneath, often the same shopper, and nearly nineteen times the wait. If you want the longer version of why Lightning behaves like that, we have been tracking its share of our Bitcoin volume for years.

The rest lands roughly where you would guess. The newer rollups, Base and Arbitrum, sit between a minute and two. Ethereum, Binance Smart Chain and Polygon cluster around two and a half minutes. Tron is three and a half, Litecoin nearly six, Dogecoin nearly seven.

The tail is where support tickets come from

A median tells you about the typical shopper. Your support inbox is filled by the other kind.

Two things stand out here, and one of them is a warning about reading medians.

More than a quarter of Bitcoin payments took longer than twenty minutes. If your team has ever wondered why the “where is my payment” emails skew so heavily towards one network, that is the answer, and it is structural rather than anything either of you did wrong.

Then Solana. The median is 66 seconds, which is excellent, and yet 3.2% of Solana payments ran past twenty minutes, a higher share than Ethereum, Tron or Polygon. Most Solana payments are close to instant and a small group is very slow. If you only looked at the median, you would never know that group existed.

The same dollar, on seven networks

This is the part that is actually actionable, because unlike the chain’s block time, it is a choice somebody is making.

USDC is one token. On our checkout it is payable on seven networks, and the wait is not the same on any two of them.

Two and a half times, top to bottom, on the same token, for the same shopper, with no difference at all in what lands in your balance.

Ether tells the same story even more sharply. Paid on Ethereum itself, the median is 2 minutes 22 seconds. Paid on Arbitrum, 1 minute 07 seconds. On Base, 1 minute 15.

However, this is the shopper’s decision, not yours. They pick the network at checkout and they pay the network fee that comes with it. What you can do is make sure the faster options are visible and not buried under whatever they picked last time.

What a merchant does with this

Four things, in rough order of how much they matter.

• Stop treating “crypto payments are instant” as a single claim. On most of our book it is nearly true. On Bitcoin it is not, and a customer paying in BTC is going to wait about a quarter of an hour on a median day.
• Write your support macros per network. A pending Bitcoin order at fifteen minutes is completely normal. A pending Solana order at fifteen minutes is not, and deserves a look.
• If you run Lightning, say so. It is the fastest thing on the list by a wide margin and it has almost no tail at all. Five orders out of 29,360 passed twenty minutes.
• Do not size your operational expectations off the median alone. Solana proves the point. A fast median and a fat tail is a real combination.

Method, and what this does not tell you

Every figure above comes from our own order data, paid orders only, paid between 1 September 2025 and 31 August 2026, pulled on 16 September 2026.

There were 1,598,666 paid orders in that window. We measured the 1,479,997 of them that carry a single identified network. The rest sit outside the tables: roughly 116,500 USDC orders where no network was recorded, and about 2,200 where the order was payable on more than one network and the record does not say which one the shopper used. Networks with fewer than 2,000 paid orders are not shown, which is why you will not see every asset we support.

Three limits worth stating plainly.

First, paid here means the network has confirmed the payment and the funds are credited to your CoinGate balance. That is what settlement actually means, and it is the point at which the payment can no longer be reversed. It is not the moment money reaches your bank, which is a separate step on a separate schedule.

Second, this is our book, not the market. Our order mix leans towards hosting, IT services and digital goods, and a merchant selling something else may see a different shopper behave differently. Our wider look at how crypto payments vary by region and asset is the closest thing we publish to context for that.

Third, the payment clock measures the interval from the shopper choosing a network to the order reaching paid. That interval contains wallet-opening time, transfer time and confirmation time together. We can separate the choice from the payment, but not the transfer from the confirmation.

Everything here is reproducible from our own data and we are happy to be checked on it.

If you are choosing which networks to put in front of your customers, the honest summary is that the modern ones are faster by roughly an order of magnitude, the difference is visible to your buyer, and it costs you nothing to enable them. Set against an international wire clearing in one to five business days, even the slowest row in that first table is a different category of wait.

Want to see which networks your own shoppers reach for? Start with us.