Why addresses look different

Open two wallets and you may well be given two addresses that look nothing alike. One begins with a 1, another with bc1q, a third with bc1p. Nothing is broken and neither wallet is wrong — but the differences are not cosmetic, and one of them will quietly cost you money every time you spend.

An address is not an account and not a location. It is a compact, human-transmissible encoding of the conditions under which these coins may later be spent. Different prefixes encode different kinds of condition.

1The four you will meet

Bitcoin has added address formats over time, always without removing the old ones. All four remain valid and spendable today.

A wallet that hands you a bc1q address is not a different kind of wallet from one handing you a 1 address. Very often it is the same seed, on the same device, walking a different branch of the same tree — which is precisely the mechanism described in how a wallet finds your coins.

2Why anyone bothered changing

Each format solved a real problem rather than being a redesign for its own sake.

P2SH, and paying to a condition

Originally, an address committed to a single public key. That made anything more sophisticated — multisig especially — awkward, because the sender had to be handed the whole complicated condition and pay for its size. P2SH inverted it: the address commits to a hash of the conditions, the sender pays for a short address regardless of complexity, and the spender reveals the full detail later. Addresses starting with 3 are the result, and it is why receiving into a multisig looks no different from any other payment.

SegWit, and moving the signatures

SegWit restructured transactions so signature data sits in a separate section, discounted when a block's size is measured. Two consequences follow: spending a SegWit output costs meaningfully less in fees, and transaction IDs stopped being malleable — a fix that Lightning depends on. Native SegWit uses bech32 encoding, which is why bc1q addresses are lowercase and slightly longer, and why they carry a checksum strong enough to catch typos rather than merely usually catching them.

Taproot, and hiding the complexity

Taproot goes further: a simple single-key spend and an elaborate multi-condition contract can look identical on-chain. If the straightforward path is taken, that is all anyone ever sees; the unused alternatives are never published. It also brought Schnorr signatures, which make the common case cheaper still.

That privacy property is not incidental — it is the whole point, and it is why complex spending conditions became practical to use without advertising to the world that you have them.

Four keys of visibly different ages laid in a row on dark cloth — an ornate old brass one through to a flat modern electronic fob — all clearly keys, all clearly for different locks.

Image to come

3The part that costs you money

Fees are charged by transaction size, not by amount sent. Address type is one of the largest influences on that size — and specifically on the size of your inputs, meaning the coins you are spending.

Approximate input sizes, which is what you pay for when spending:

• Legacy (1…) — around 148 vbytes per input. The most expensive by a wide margin.
• Wrapped SegWit (3…) — around 91 vbytes.
• Native SegWit (bc1q…) — around 68 vbytes.
• Taproot (bc1p…) — around 58 vbytes for an ordinary single-key spend.

Spending a single coin, the difference is small change. Consolidating twenty small legacy inputs versus twenty Taproot ones is roughly 1,800 vbytes of difference — a real sum during a busy fee period, and the reason fees and address types are the same conversation.

4Compatibility, and the shrinking list of exceptions

Adoption lags activation by years, because every service has to update its own software.

• Some older exchanges and services still cannot send to Taproot (bc1p) addresses, and a smaller number still struggle with bc1q. If a withdrawal form rejects your address, this is usually why.
• The fix is to receive to a format the sender supports — most wallets can produce all four from the same seed — rather than to abandon the withdrawal.
• Sending to an old-format address always works. The limitation is only ever on the sending side.

Bech32 addresses are case-insensitive, so bc1q… and BC1Q… are the same address. They are conventionally shown in lowercase, and a QR code of one may be uppercase purely because that encodes more compactly. Neither is a sign of anything wrong.

5Practical guidance

• Take your wallet's default unless you have a reason not to. Modern wallets default to native SegWit or Taproot, both of which are fine.
• Use a fresh address for every payment. Reusing one links payments together permanently and publicly — see the privacy you actually have. This matters far more than which format you picked.
• Do not chase the newest format for its own sake. The saving between bc1q and bc1p is a few percent of an already small number for most people.
• Do note the format alongside your backup, because it determines the derivation path a future restore has to use.
• Verify the address on your signing device's screen, whatever the format. Address-swapping malware does not care which prefix you use.

The short version

Four formats, all valid, all spendable, differing in what they commit to and what they cost to spend. Newer ones are cheaper and better at hiding complexity; older ones are more widely accepted. Your wallet's default is almost certainly the right answer, and the format you use matters considerably less than not reusing whichever one you chose.