Ordinals and Inscriptions

In January 2023, an image was etched onto the Bitcoin blockchain. It was a small piece of pixel art, but the event threw the Bitcoin community into one of its most heated debates since the Blocksize War. Some hailed it as Bitcoin’s renaissance; others condemned it as spam that defiled the protocol’s sacred purpose. At the center of this storm are Ordinals and Inscriptions.

The idea that Bitcoin could serve as a permanent data medium beyond simple value transfer, and that individual satoshis could carry unique identities, challenges fundamental assumptions about what Bitcoin is and what it should become. These are not merely technical questions. They are philosophical ones that strike at the heart of Bitcoin’s social contract.

What Is Ordinal Theory?

Ordinal theory is a system proposed by software developer Casey Rodarmor in January 2023 that assigns unique serial numbers to individual satoshis, the smallest unit of bitcoin (1 BTC = 100,000,000 satoshis).

The Bitcoin protocol itself does not distinguish between satoshis. One satoshi in your wallet is identical to one satoshi in anyone else’s. This property is called fungibility, and it is a cornerstone of money. A dollar bill is a dollar bill regardless of its serial number.

What Rodarmor demonstrated is that it is possible to artificially impose an ordering on these fungible units. By numbering each satoshi according to the order in which it was mined, and tracking those numbers through transactions using a first-in-first-out (FIFO) rule, every satoshi can be individually identified.

The name “ordinal” comes from mathematics. An ordinal number indicates position in a sequence: first, second, third. Ordinal theory applies this sequential numbering to satoshis.

The Numbering Scheme

The rules for assigning numbers to satoshis are remarkably simple.

Creation rule: When a new block is mined, the satoshis in the coinbase reward are numbered sequentially. The genesis block (block 0) produced satoshis numbered 0 through 4,999,999,999. Block 1 produced satoshis starting at 5,000,000,000, and so on.

Transfer rule: When a transaction spends inputs and creates outputs, the satoshis from inputs flow into outputs in first-in-first-out order. The satoshis from the first input fill the first output, then the second output, and so on. Once the first input’s satoshis are exhausted, the second input’s satoshis continue the sequence.

The critical insight is that this numbering does not happen inside the Bitcoin protocol. Ordinal theory is an interpretive lens applied to existing blockchain data. It requires no protocol changes. It is a social layer, a convention, a shared agreement to treat satoshis as if they have individual identities.

Inscriptions: Writing Data onto Bitcoin

If ordinal theory gives satoshis identity, inscriptions are the mechanism for permanently attaching data to those identified satoshis.

The core idea is to embed arbitrary data — images, text, audio, video, HTML, even executable code — into the witness data section of a Bitcoin transaction. This data is permanently recorded on the Bitcoin blockchain and cannot be deleted or altered by anyone, as long as the Bitcoin network exists.

Technically, inscriptions work as follows:

1. A special transaction is created that includes data within a Taproot script.
2. The data is wrapped in an OP_FALSE OP_IF ... OP_ENDIF structure. This creates dead code that does not affect the transaction’s validity but is permanently recorded on the blockchain.
3. The data envelope includes a content type (MIME type) and the actual content bytes.
4. When this transaction is included in a block, the data is considered “inscribed” onto a specific satoshi.

The result: a particular satoshi gains both a unique number and attached data. This is, effectively, an NFT on the Bitcoin blockchain.

Why SegWit and Taproot Made This Possible

Inscriptions became feasible because of two major Bitcoin protocol upgrades that were designed for entirely different purposes.

SegWit (2017)

Segregated Witness, activated in 2017, separated transaction signature data into a separate area called the witness. The crucial detail is that witness data receives a 75% fee discount compared to regular transaction data. One byte of witness data costs only 0.25 weight units, versus 1 weight unit for non-witness data.

This “weight unit” system effectively expanded the block size limit from 1MB to a theoretical maximum of 4MB, but only if the block consists almost entirely of witness data.

Taproot (2021)

Taproot, activated in November 2021, removed the practical size limit on scripts that can be placed in the witness area. Before Taproot, individual script elements were limited to 520 bytes. Taproot’s tapscript eliminated this constraint.

Combined, these two upgrades create the perfect conditions for inscriptions: Taproot allows large data payloads in the witness area, and SegWit’s discount structure makes storing that data relatively cheap. Inscriptions exploit this unintended synergy.

Neither SegWit nor Taproot was designed with blockchain-based image storage in mind. SegWit aimed to fix transaction malleability and improve scalability. Taproot aimed to enhance privacy and smart contract flexibility. The fact that inscriptions emerged from these upgrades is a case study in how complex systems produce emergent behaviors that their designers never anticipated.

BRC-20 Tokens

Once inscriptions proved viable, an anonymous developer known as @domodata pushed the concept further in March 2023. Using JSON-formatted inscriptions, they proposed a standard for issuing fungible tokens on Bitcoin. This became BRC-20.

The name is borrowed from Ethereum’s ERC-20 token standard, but the mechanism is fundamentally different. Ethereum’s ERC-20 uses smart contracts to manage token balances and transfers on-chain. BRC-20 simply records JSON text as inscriptions.

{
  "p": "brc-20",
  "op": "deploy",
  "tick": "ordi",
  "max": "21000000",
  "lim": "1000"
}

The above inscription deploys a token called “ordi” with a maximum supply of 21 million and a per-mint limit of 1,000. Users then create “mint” inscriptions to claim tokens and “transfer” inscriptions to send them.

The defining characteristic — and limitation — of BRC-20 is its dependence on off-chain indexers. The Bitcoin protocol does not interpret this JSON. Separate indexer software scans the blockchain, reads BRC-20 inscriptions, and computes token balances. Different indexers could theoretically reach different conclusions, making the consensus mechanism less robust than Bitcoin itself.

Despite this limitation, BRC-20 exploded in popularity. In May 2023, the majority of Bitcoin network transactions were BRC-20-related, driving fees to levels that priced out ordinary transfers. It was a real-time experiment in what happens when new sources of demand compete for limited block space.

The Bitcoin Community Debate

Ordinals and inscriptions triggered one of the fiercest debates in Bitcoin’s history, rivaling the Blocksize War of 2015-2017 in its intensity and ideological depth.

Arguments in Favor

New utility and ecosystem growth: Ordinals give Bitcoin value beyond money. Digital art, collectibles, gaming items, and domain names can now exist on the most secure blockchain in the world. This attracts new users and developers to the Bitcoin ecosystem who might otherwise have gravitated toward alternative chains.

Increased miner fee revenue: Bitcoin’s block subsidy halves approximately every four years. In the long run, miner income must come from transaction fees. Inscriptions have dramatically increased demand for block space. On several occasions since 2023, daily fee revenue has exceeded the block subsidy — a positive signal for Bitcoin’s long-term security budget.

Free market for block space: Bitcoin is a permissionless network. Anyone who pays the fee can use block space. Deciding which uses are “legitimate” contradicts Bitcoin’s core philosophy. The market sets the price, and users compete freely. If someone is willing to pay $50 to inscribe an image, that is the market speaking.

On-chain permanence: Unlike many Ethereum NFTs that store images on external servers (IPFS or centralized hosts), Bitcoin inscriptions contain the data itself on-chain. As long as a single Bitcoin node exists, the data persists. There are no broken links, no disappeared images, no rug-pulled metadata.

Arguments Against

Block space waste: Bitcoin’s block space is an extraordinarily scarce resource. With a capacity of roughly 7 transactions per second, using this space to store JPEG images is wasteful. This space should serve the economic freedom of billions — value transfer, not data storage.

Fee escalation: Competition for block space from inscriptions drives up fees for everyone. During May 2023, when average transaction fees exceeded $30, small-value users in El Salvador and other developing countries were effectively priced out of the network. Bitcoin’s promise as “a bank for the unbanked” is undermined when a remittance costs more than the amount being sent.

Fungibility erosion: Assigning numbers to satoshis and claiming some are more “valuable” than others undermines Bitcoin’s fungibility. For Bitcoin to function as money, 1 sat must equal 1 sat. The existence of “special” satoshis erodes this principle and could eventually create regulatory complications if certain satoshis are deemed to carry tainted data.

Identity crisis: Is Bitcoin sound money or a general-purpose data store? Embracing ordinals dilutes Bitcoin’s core narrative. Bitcoin should be a protocol that does one thing exceptionally well, following the Unix philosophy of simplicity. Adding cultural layers of NFTs and tokens creates noise that distracts from the monetary mission.

OP_RETURN vs Inscriptions

Recording data on the Bitcoin blockchain is not new. Since 2014, the OP_RETURN opcode has allowed small amounts of data to be attached to transactions.

Property	OP_RETURN	Inscriptions
Maximum size	80 bytes	~4MB (block weight limit)
Storage location	Output script	Witness data
UTXO set impact	None (prunable)	None (witness area)
Fee weight	1x (standard cost)	0.25x (witness discount)
Typical use	Timestamps, hash commitments	Images, text, media

OP_RETURN was intentionally limited to 80 bytes to constrain non-monetary blockchain usage. This was a deliberate design decision to prevent the blockchain from becoming a data storage platform. Inscriptions circumvent this philosophy by leveraging the witness discount and Taproot’s relaxed script size limits to store megabytes of data.

It is worth noting that OP_RETURN itself was controversial when introduced. The same argument — “Bitcoin is not a data store” — was made in 2014. The pragmatic conclusion then was that people would embed data regardless, so providing a method that does not bloat the UTXO set was the lesser evil. The ordinals debate is, in many ways, a replay of this same tension at a much larger scale.

Rare Satoshis (Rare Sats)

Ordinal theory goes beyond simple numbering to classify certain satoshis into rarity tiers. This mirrors the world of coin collecting, where coins from specific years or with minting errors command premiums.

• Common: Any satoshi that is not the first in its block
• Uncommon: The first satoshi of each block (~6,929,999 possible)
• Rare: The first satoshi of each difficulty adjustment period (~3,437 possible)
• Epic: The first satoshi of each halving epoch (4 so far, maximum 32)
• Legendary: The first satoshi of a block where difficulty adjustment and halving coincide (theoretically possible, none yet)
• Mythic: The first satoshi of the genesis block (exactly 1, but locked in an unspendable coinbase)

This rarity system is entirely based on social consensus. The Bitcoin protocol has no awareness of these classifications. Yet a collector market has formed, with “epic” satoshis trading for thousands of dollars. This is a fascinating social experiment that reveals how humans instinctively assign value to scarcity, even when that scarcity is based purely on convention.

Bitcoin NFTs vs Ethereum NFTs

Inscriptions are often called “Bitcoin NFTs,” but they differ from Ethereum-based NFTs in fundamental ways.

Data Storage

The majority of Ethereum NFTs use off-chain metadata. The token contract stores only a URL pointing to the image, which is hosted on IPFS or a centralized server. If that server goes down, the NFT’s image disappears. On-chain art like Autoglyphs exists on Ethereum, but storing large data on-chain is prohibitively expensive due to gas costs.

Bitcoin inscriptions contain the complete data on the blockchain. Every byte of the image is recorded in a Bitcoin block. As long as one Bitcoin node exists anywhere in the world, the data survives.

Ownership Model

Ethereum NFT ownership is managed by smart contract state. Trading happens through specialized marketplaces (OpenSea, Blur) and complex contract calls with approval mechanisms.

Bitcoin inscription ownership is tied to the satoshi itself — specifically, to whoever controls the UTXO containing that satoshi. Transfer the satoshi, and the inscription moves with it. This model is simpler but carries the risk of accidentally spending a “valuable” satoshi as transaction fees.

Smart Contract Functionality

Ethereum NFTs can incorporate automatic royalty payments, auctions, fractional ownership, and complex programmable logic. Bitcoin inscriptions have none of this. They are simply data attached to satoshis with no automated logic. Some view this as a feature: less complexity means a smaller attack surface and fewer exploits.

Impact and the Future

Structural Changes to the Fee Market

Inscriptions have permanently altered Bitcoin’s fee market dynamics. Before 2023, fees were driven primarily by network congestion during price rallies. Now, inscription demand creates a new baseline for fees. Popular inscription project mints can spike fees dramatically, directly affecting regular transaction users.

In the long term, this could be part of the answer to Bitcoin’s security budget question. As the block subsidy approaches zero over the coming decades, can fees alone provide sufficient incentive for miners to secure the network? Inscriptions create a new source of demand for block space, contributing to fee revenue that helps sustain network security.

Cultural Shift

Ordinals have added a new cultural dimension to Bitcoin. “Sat hunters” — collectors who track and trade rare satoshis — have emerged as a distinct subculture. A Bitcoin art scene has developed, with artists choosing to inscribe their work on the most secure and permanent blockchain in existence.

Whether one celebrates or laments this shift, it has broadened the entry points to Bitcoin. Someone who has no interest in monetary theory might discover Bitcoin through digital art, and from there learn about sound money, self-custody, and financial sovereignty.

Relationship with Layer 2

The relationship between inscriptions and Bitcoin’s Layer 2 solutions, particularly the Lightning Network, is complex. Higher base-layer fees driven by inscriptions increase the cost of opening Lightning channels, but simultaneously reinforce the case for why Layer 2 scaling is essential.

Some argue that ordinals functionality should migrate to Layer 2. The base layer should remain an efficient settlement layer for high-value transactions, while data storage and token issuance belong on higher layers. Projects like Stacks, RGB, and Liquid are building programmable layers on top of Bitcoin that could absorb this demand.

The Censorship Debate

Some Bitcoin miners and node operators have attempted to filter inscription transactions. Alternative node software like Bitcoin Knots classifies inscriptions as “spam” and offers options to refuse relaying them.

This approach collides directly with Bitcoin’s core value of censorship resistance. If Bitcoin begins discriminating between types of valid transactions, it ceases to be a truly censorship-resistant network. A valid transaction is a valid transaction, whether it transfers value or records data. The moment subjective judgments about “acceptable” uses enter the protocol, Bitcoin loses its most powerful property.

The market is providing its own answer. Inscription transactions pay fees and comply with all rules of valid Bitcoin transactions. The price mechanism of block space allocates resources, and this market-based approach is gradually recognized as a more Bitcoin-native solution than censorship.

Conclusion

Ordinals and inscriptions have demonstrated that Bitcoin remains a living protocol capable of evolving in unexpected directions. When Satoshi Nakamoto designed Bitcoin in 2009, it is unlikely they imagined that fourteen years later people would be etching digital art onto the blockchain.

Whether this phenomenon is ultimately beneficial or harmful to Bitcoin remains an open question. What is certain is that competition for the scarce resource of block space is making Bitcoin’s economic model more sophisticated, and the debate itself demonstrates the health of Bitcoin’s governance. In a centralized system, an administrator would decide whether to permit inscriptions. In Bitcoin, tens of thousands of node operators, miners, developers, and users each make their own choices, and the network’s direction emerges from their collective decisions.

Bitcoin is not controlled by any single vision. It evolves through code, economic incentives, and the consensus of its participants. Ordinals are the most recent — and most dramatic — chapter in that ongoing evolution.