Ever notice how Bitcoin keeps surprising people? It’s not just digital gold anymore. Ordinals and inscriptions turned sats into carriers of arbitrary data, and suddenly you can mint what looks and feels like an NFT right on Bitcoin. This caught a lot of folks off guard—some loved it, others said it was messy. I’m going to walk through how ordinals work, why inscriptions matter, and what BRC-20 tokens really are, without the fluff.

Short version: Ordinals are a numbering system for individual satoshis (the smallest Bitcoin unit). Inscriptions attach data—images, text, scripts—to those numbered sats using the witness portion of a Taproot transaction. That makes the data part of Bitcoin’s chain history without touching legacy script space. The design is clever, and practical in ways people didn’t predict.

What is an Ordinal? The core concept

At the base level, an ordinal assigns a serial number to each satoshi by tracking issuance order. That sounds academic, but it gives you a way to point to a specific satoshi inside a UTXO. Once you can reference an exact satoshi, you can attach metadata to it. That’s the opening move that makes inscriptions and on-chain artifacts possible.

Technically, ordinals use transaction order and block data to identify sats. No changes to Bitcoin protocol were made; the mechanism is out-of-band interpretation. The community layer reads a sat’s path through transactions to identify it. Because it’s interpretive, wallets and explorers must adopt ordinal-aware logic to surface these inscriptions to users.

Inscriptions: how data gets written

Inscriptions store arbitrary data (like an image or a small program) in a transaction’s witness field, typically in a Taproot output. That’s important: witness data is segregated from legacy scriptSig space and was introduced by SegWit and extended by Taproot. The result is that developers can embed reasonably large blobs of data without violating the non-standard script limits of older transaction formats.

When an inscription is attached to a sat, the ordinal protocol treats that sat as “the holder” of the inscription. Transfer the sat—transfer the inscription. It’s literal. Readability depends entirely on tooling: explorers, wallets, and marketplaces must decode the witness and present the content. If they don’t, the inscription exists on-chain but stays invisible to most users.

BRC-20: token standards on Bitcoin, sort of

Okay, BRC-20—this one gets a lot of hype. BRC-20 is an experimental token standard built on inscriptions. It mimics the spirit of ERC-20 (Ethereum) but is very different technically. A BRC-20 token is a convention: JSON data inscribed in particular ways that wallets and indexers parse to create fungible token balances off-chain (tracked by ordinal-aware indexers).

Important caveats: BRC-20 tokens aren’t native smart contracts. There’s no consensus-level token accounting like Ethereum’s contract state. Instead, indexers read the history of inscriptions and infer mint, transfer, and burn actions. That approach is flexible and hacky—cool for experiments, but it has limitations around atomicity, composability, and standard security guarantees you get from EVM contracts.

Why people like Ordinals—and why some are worried

Pros: Bitcoin gains an expressive layer without a soft fork. Creators can place digital art, memes, or small programs directly on-chain, benefiting from Bitcoin’s immutability and wide distribution. Collectors enjoy “on-chain everything” authenticity. For many artists and developers, that’s valuable.

Cons: Fees and block space. Inscribing large files means bigger witness data and higher fees, which can spike during busy periods. Some community members also felt it clashed with Bitcoin’s priorities—store-of-value advocates worry about bloat, node operators worry about storage costs, and others point out that inscriptions change usage patterns for UTXO set and indexing.

Practical steps to inscribe or collect safely

If you want to view or hold ordinals, use ordinal-aware wallets and explorers. For managing inscriptions, sign with tools that explicitly support inscriptions and verify outputs yourself. Always remember: inscriptions are immutable and live on-chain forever; mistakes are permanent.

One popular tool in the ecosystem is the unisat wallet, which many users rely on to inspect and transact with ordinals. It provides an interface to see inscriptions and manage ordinal sats, though every tool has trade-offs—review permissions, test with small amounts first, and keep your recovery seeds offline and safe.

Costs, fees, and performance

Inscribing big files costs more. Because inscription data is included in the transaction witness, fee calculation depends on virtual size (vbytes) and mempool demand. Expect to pay market-rate fees; if blocks are full, fees climb. Pro tip: compress content, use efficient formats, and consider storing only a content fingerprint on-chain with a compressed data model if you care about cost.

Node operators: a dense inscription history can increase indexing and storage requirements. Full nodes still validate everything, but optional indexers that surface ordinals need more careful resource planning. If you run services around ordinals, monitor disk usage and plan for growth.

Best practices for creators and collectors

Creators should think about intent: do you want the full asset on-chain, or is an on-chain reference enough? If authenticity and permanence are crucial, a full inscription makes sense. But if cost-efficiency matters, hybrid approaches (small on-chain proof + off-chain hosting) are reasonable.

Collectors: verify provenance. Because inscriptions are public and permanent, tracing the transaction history is possible—but corroborate with trusted indexers or explorer tools. Use ordinal-aware wallets that let you inspect the raw inscription and its transaction metadata.