Okay — so you’ve seen the headlines: “Bitcoin NFTs” and “Ordinals” popping up everywhere. They feel new, and a little weird. Really? Bitcoin doing what Ethereum did? Sort of, though the mechanics are different. This piece walks through how wallets handle ordinals inscriptions, what that means for BRC-20 tokens and “Bitcoin NFTs,” and practical tips for anyone who wants to try it without burning sats or losing their keys.

First, the quick orientation. Ordinals are a way to index individual satoshis (the smallest units of BTC) so you can attach arbitrary data to them via an on‑chain transaction. Those attachments are called inscriptions. They’re permanent and live on Bitcoin’s UTXO set forever once confirmed. That permanence is both the appeal and the problem — it’s irreversible and on-chain storage costs real fees.

Wallet fundamentals for Ordinals and inscriptions

Not every wallet supports ordinals. Many traditional wallets treat Bitcoin as fungible, which it largely is for payments. But ordinals require three practical wallet features: precise UTXO control, the ability to craft custom transactions (including data-carrying outputs via Taproot or OP_RETURN where applicable), and visibility into which satoshis are inscribed. If your wallet hides coin-level details, you can accidentally spend an inscribed sat — poof, your inscription may move or be destroyed in unexpected ways.

One wallet that became popular among ordinal users offers a straightforward UX for managing inscriptions — the unisat wallet — it shows inscriptions tied to specific sats and lets you create or transfer them with more control than typical SPV wallets. I’m mentioning it because it illustrates the features you should look for: clear UTXO view, explicit inscription send flow, hardware-wallet compatibility.

Note: “BRC-20” tokens are a bit different. They piggyback on the ordinal paradigm to create fungible-like tokens using inscriptions as state transitions. They’re not smart contracts in the Ethereum sense. Instead, they rely on off-chain tooling and conventions that read inscriptions to infer token balances. So your wallet needs to support whatever indexer and parsing logic is used by the BRC-20 ecosystem to display balances correctly.

How an inscription actually gets on-chain — step by step

Here’s the short version. You pick a UTXO (a set of sats). You create a transaction where one output carries your payload (the inscription) encoded according to the ordinal standard, and the rest goes back as change. Pay fees. Broadcast. Confirm. The network writes the data into the UTXO and it becomes discoverable by indexers and ordinal-aware explorers.

Digging a little deeper: inscriptions today typically use Taproot outputs to store the payload in witness data, which is cheaper than storing in legacy script fields and avoids some policy limits. Wallets that craft inscriptions must construct the transaction carefully: choose a UTXO with enough sats to cover the output plus fees, ensure the data is formatted per the ordinal spec, and set a fee rate that miners will accept (inscriptions can be large, so they often need higher absolute fees).

Price note: size matters. A 1KB inscription costs more than a 10B one. During high demand, fees spike — that’s when inscription costs can be very high. Plan ahead and estimate fee in sat/vByte terms.

Security and UX: things that matter

Don’t get sloppy. If a wallet doesn’t support PSBT or hardware signing, don’t use it for high-value inscriptions. Use a hardware wallet when possible. Also: back up your seed properly. Inscribed sats are still sats; losing the seed means losing access. That part is basic, but worth repeating.

Watch how coin selection works. Some wallets automatically combine many UTXOs into a single transaction. That can unintentionally consolidate inscribed sats with others and make later transfers messy or expensive. Wallets that let you manually pick UTXOs save fuel — and grief.

Another operational tip: watch mempool behavior. Long confirmation times can lead to stuck inscriptions. If your wallet supports bumping fees (RBF) or replacement logic, that’s extremely helpful. If not, be prepared to wait or reconstruct carefully.

Practical workflows for creators and collectors

Creators — if you plan to inscribe media (images, JSON metadata, short videos): consider off-chain hosting with on-chain pointers versus full on-chain data. Full on-chain is elegant and immutable, but expensive. Many creators write a small JSON manifest on-chain that points to a content hash hosted elsewhere (IPFS). That keeps costs down while preserving verifiability.

Collectors — verify inscriptions with multiple explorers. Some indexers display different metadata. If you’re buying from a marketplace or P2P, check the exact satoshi identifiers and see them on-chain. And again: escrow or smart intermediaries aren’t native to Bitcoin the way they are on some other chains, so trade carefully.

Common pitfalls and how to avoid them

Sending the wrong UTXO. Easy to do. Always confirm which sat (or sat group) you’re sending.

Underpaying fees. Worse when inscription sizes are large. Use real-time fee estimates.

Using custodial services that don’t surface ordinals. They might ignore or drop inscriptions entirely. If the inscription matters, custody yourself or use a provider that explicitly supports ordinals.

Tooling and ecosystem notes

Indexers and explorers are the lenses through which ordinals are seen. Some are community-run and others are commercial. They parse witness data, map inscriptions to sat indices, and build galleries. Wallets integrate with these indexers to present inscriptions in a friendly UI. If you care about long-term discoverability, prefer wallets that use reputable indexers and allow exportable proofs or raw transaction views.

BRC-20 token support depends on conventions not protocol-level features. That means fragmentation: different wallets and marketplaces may interpret inscriptions differently. Keep that in mind before assuming interoperability.