Buying Gift Cards Over Lightning: When it Beats On-Chain Bitcoin

Summary: An on-chain Bitcoin fee is priced by transaction size, not by how much you are sending, so it lands hardest on small purchases. Lightning prices by amount instead, which is why a routing fee on a modest gift card is usually a fraction of a cent against a dollar or more on-chain. It settles in seconds rather than tens of minutes. The trade is that Lightning has a size ceiling, the invoice expires, and your wallet has to support it.

• On-chain fees are charged per byte of transaction data, so a small purchase pays roughly the same as a large one.
• Lightning routing fees are typically a tiny percentage of the amount, often well under a cent on everyday sums.
• Reliability inverts above a few hundred dollars, because a payment cannot exceed the liquidity of the smallest channel on its route.
• A Lightning invoice expires. A Bitcoin address does not. Treating one like the other is the most common mistake.

Lightning is available for Bitcoin and Litecoin payments, though the argument below is really a Bitcoin argument. It sits alongside the other routes for spending bitcoin, and for a small Bitcoin gift card purchase it is often the cheaper rail by a wide margin. It is not the right rail for every purchase, and the cases where it fails are predictable enough to plan around.

What Lightning Actually Changes

Lightning sits on top of Bitcoin rather than replacing it. Payments move through channels between participants and settle off the main chain, so they are not waiting on a block. The base layer stays the security and settlement layer, and Lightning becomes the layer for the small, frequent payments the base layer was never designed to carry.

Two practical differences follow. Settlement is fast, with production measurements in 2026 putting typical payments around one to two seconds and independent withdrawal tests landing between roughly 43 and 52 seconds, against about 14 minutes for a standard on-chain withdrawal. And the cost structure is completely different, which is the part that matters most for a gift card.

Why On-Chain Punishes Small Purchases

An on-chain fee is calculated from how much block space your transaction occupies, measured in virtual bytes. It has almost nothing to do with the value being sent. A transaction moving 20 dollars and a transaction moving 20,000 dollars can cost the same to confirm.

That produces an uncomfortable result at the low end. During congestion an on-chain transaction can run anywhere from a dollar to well over ten, and at the extremes has reached the tens of dollars. Send five dollars of Bitcoin on-chain at the wrong moment and the fee can exceed the payment. For a 25 dollar gift card that is a meaningful share of what you are spending.

Lightning inverts the relationship. Each hop charges a small base fee plus a proportional rate measured in parts per million, so the cost scales with the amount rather than the data. Published benchmarks put typical routing fees between roughly 0.001 and 0.05 percent of the payment. On a payment worth around 75 dollars that is a few cents at most, and often a fraction of one.

Where Lightning Wins

The pattern is consistent: the smaller and more ordinary the purchase, the bigger the advantage.

• Small denominations. A 10 or 25 dollar card is exactly the case where a fixed on-chain fee hurts most.
• Buying several cards over time. The saving compounds where a single purchase might not justify the setup.
• When you want the code quickly. Seconds rather than waiting on confirmations.
• When the mempool is busy. Lightning costs barely move while on-chain fees are spiking.

Reported routing reliability supports this. Success rates sit around 95 to 99 percent for payments under roughly 500,000 satoshis, which covers essentially every normal gift card purchase.

Where Lightning Loses

Being honest about the ceiling matters more than the sales pitch, because this is where people get stuck.

A Lightning payment travels a route made of channels, and it cannot exceed the capacity on the smaller side of any channel along that route. Median channel size on the public network sits near 0.02 BTC, so a large payment can fail at a single hop even when the network holds plenty of aggregate liquidity and both wallets show a healthy balance.

The reliability numbers reflect that. Success rates fall to roughly 80 to 90 percent in the range of one to ten million satoshis and drop off sharply above that. So for a large purchase, on-chain is the more predictable choice, and the extra fee stops being significant once the amount is big enough. The rule of thumb is that the fixed cost of on-chain matters less the more you are sending.

The Invoice That Expires

This catches more people than the liquidity ceiling. A Bitcoin address is permanent and can be reused. A standard Lightning invoice is single-use by specification and carries an expiry that can be anything from a few minutes to a few hours. Sit on the checkout page too long and the invoice stops being valid, and you need a fresh one.

Practically, that means generating the invoice when you are ready to pay rather than opening it and going to find your wallet. If it lapses, nothing is lost. Start again and a new invoice appears. Newer reusable payment requests are being adopted and remove this friction, but single-use invoices are still what you will meet most often.

What You Need on Your Side

Support is not universal, so check before you commit to the route.

• A Lightning-capable wallet, which by definition means a hot wallet, since it has to be online to route a payment. Not every Bitcoin wallet speaks Lightning, and an on-chain-only wallet cannot pay an invoice.
• A balance already on Lightning, or a wallet that moves funds across for you.
• No node of your own. Modern wallets handle channels and routing automatically, and running a node is an advanced option rather than a requirement.

Litecoin is the honest exception. The fee problem this article describes is a Bitcoin problem, and Litecoin's on-chain fees are already small enough that Lightning solves very little there. The Litecoin Lightning network is also tiny: the Litecoin Foundation's own write-up counted 95 nodes, 170 channels and about 35 LTC of capacity in May 2024, and concluded that running Lightning on Litecoin does not currently make much sense. Almost every consumer Lightning wallet is Bitcoin-only. If you are paying in Litecoin, on-chain is usually the simpler choice.

When to Reach for Lightning

Pick the rail that matches the shape of the payment rather than the one described as faster. Small and everyday goes over Lightning, where the fee is negligible and the code arrives in seconds. Large and infrequent goes on-chain, where the fixed fee is a rounding error and the routing ceiling is not a factor.

If you are unsure whether your coin and network are covered, the supported currencies list is the place to check before starting a purchase.