In February 2023, Bitcoin was trading at $24,641, recovering from its first red week of the year, and the Ordinals protocol had just surpassed 130,000 inscriptions since its January launch. The Bitcoin network — long considered too rigid for anything beyond simple value transfers — was suddenly hosting a thriving NFT ecosystem. But how exactly do Ordinals work under the hood? Unlike Ethereum’s ERC-721 tokens, which rely on smart contracts to manage ownership and metadata, Ordinals take a fundamentally different approach: they embed data directly into Bitcoin transactions, leveraging the network’s native UTXO model and Taproot upgrade to create unique, immutable digital artifacts. This tutorial walks through the technical architecture, inscription process, and verification mechanics of Bitcoin Ordinals.
The Objective
This guide aims to provide a comprehensive technical understanding of how the Ordinals protocol enables NFT-like functionality on Bitcoin without modifying the blockchain’s consensus rules. By the end, you should understand the theoretical foundations (ordinal theory), the practical inscription process (how data is embedded in transactions), and the verification mechanics (how inscriptions are tracked and validated).
The Ordinals protocol, created by developer Casey Rodarmor, is built on a concept called ordinal theory — a system for individually identifying and tracking every satoshi (the smallest unit of Bitcoin, equal to 0.00000001 BTC) on the blockchain. While Bitcoin transactions have always involved individual satoshis, the protocol had no native way to distinguish one satoshi from another. Ordinal theory assigns a unique number to each satoshi based on the order in which it was mined, creating a persistent identity for every unit of Bitcoin.
Prerequisites
Before diving into the technical details, you should have a working understanding of the following concepts: Bitcoin’s UTXO (Unspent Transaction Output) model, where transactions consume previous outputs and create new ones; the Taproot soft fork (activated November 2021), which introduced Schnorr signatures and Tapscript; and basic familiarity with Bitcoin transaction structure, including inputs, outputs, and witness data.
You will also need a Bitcoin node running in pruned or full mode with the Ordinals indexer software, which tracks the location of every satoshi and maintains a database of inscriptions. The ord client is a Rust-based command-line tool that communicates with your Bitcoin node to create, track, and verify inscriptions.
Step-by-Step Walkthrough
Step 1: Understanding Ordinal Numbers. Every satoshi mined receives a sequential ordinal number based on its position in the blockchain. The first satoshi in the genesis block is ordinal 0, the second is ordinal 1, and so on. As of February 2023, approximately 19.37 million BTC (nearly 1.937 quadrillion satoshis) had been mined. The ordinal numbering scheme follows a specific hierarchy: satoshis are numbered within blocks, blocks within difficulty periods, difficulty periods within halving epochs, and halving epochs within the total supply.
Step 2: The Inscription Transaction Structure. An inscription is created through a specific Bitcoin transaction pattern that embeds arbitrary data into the witness field of a Taproot input. The process involves two key components: a commitment transaction and a reveal transaction. The commitment transaction creates a Taproot output that contains the inscription data in its script path spend condition but does not expose the data on-chain. The reveal transaction spends this output, making the inscription data visible in the blockchain.
The inscription data is encoded as an envelope in the Tapscript using OP_FALSE OP_IF … OP_ENDIF opcodes. This envelope structure includes content type headers (such as image/png or text/html) followed by the actual data payload. Because the data is wrapped in a conditional that always evaluates to false (OP_FALSE), the data is never actually executed by the Bitcoin script interpreter — it is simply stored in the witness data.
Step 3: Transferring Inscriptions. Once created, an inscription is permanently associated with the specific satoshi that was inscribed. Transferring the inscription is simply a matter of transferring that satoshi in a Bitcoin transaction. The Ordinals indexer tracks the location of each ordinal-numbered satoshi as it moves through transaction inputs and outputs, maintaining a continuous chain of custody for every inscription.
This transfer mechanism is entirely native to Bitcoin — no special protocols, sidechains, or additional consensus rules are required. Any Bitcoin wallet can technically transfer an inscribed satoshi, though using the ord client ensures that the inscription is preserved during the transfer (i.e., the inscribed satoshi is not accidentally split or combined with other satoshis).
Step 4: Verification and Indexing. To verify an inscription, you need to trace the ordinal number of its satoshi back through the blockchain to the original inscription transaction. The ord indexer maintains a local database that maps ordinal numbers to their current UTXO locations and tracks all inscription metadata. When you query the indexer for a specific inscription, it returns the content type, data payload, current owner (as the Bitcoin address holding the UTXO), and the complete history of transfers.
Troubleshooting
The most common issue encountered when working with Ordinals is accidentally spending an inscribed satoshi as a transaction fee. Because Bitcoin wallets typically select UTXOs without regard for their ordinal numbers, a standard wallet transaction might consume an inscribed satoshi as an input and then allocate it to the miner fee rather than a specific output. Using the ord wallet, which is aware of ordinal numbers, prevents this by ensuring that inscribed satoshis are always sent to a designated output.
Another frequent issue is excessive transaction fees during periods of high inscription activity. In February 2023, the popularity of Ordinals drove Bitcoin transaction fees higher as inscription transactions competed for block space. A single inscription transaction with a large image payload could consume significant block space, leading to higher fees. Optimizing image sizes and batching inscriptions can help manage costs.
Indexing performance can also be a bottleneck. The ord indexer must process every Bitcoin transaction to track ordinal numbers, which requires significant computational resources and storage. As of February 2023 with 130,000+ inscriptions and 800,000+ blocks, the full index required several hundred gigabytes of storage and could take days to build from scratch.
Mastering the Skill
Bitcoin Ordinals represent a paradigm shift in how we think about the Bitcoin blockchain’s capabilities. By leveraging existing consensus features — particularly Taproot and the witness data field — the protocol demonstrates that Bitcoin can support complex data storage and NFT functionality without any protocol-level changes. As the ecosystem matures, expect to see more sophisticated inscription formats, improved indexing tools, and specialized wallets designed for managing ordinal-based digital artifacts. The 130,000 inscriptions milestone reached in February 2023 was just the beginning — the true potential of Bitcoin-native digital artifacts is still being explored.
Disclaimer: This article is for educational purposes only and does not constitute financial advice. Always conduct your own research before interacting with cryptocurrency protocols.
130k inscriptions and the tech is barely a month old. the taproot angle is clever – no consensus changes needed. this is how you build on bitcoin without begging miners for a fork
taproot enabling this is ironic given it was marketed as a privacy upgrade. now the blockchain is full of jpeg data lol
the irony of taproot enabling NFTs when it was sold as a privacy upgrade is peak bitcoin governance. unintended consequences everywhere
jpeg_miner_ taproot was sold as a privacy upgrade and the first thing people did was put JPEGs on chain. most bitcoin governance moment possible
taproot was marketed as privacy and the first major use case was putting JPEGs on chain. bitcoin governance in a nutshell
the UTXO model is what makes ordinals work so cleanly. each inscription is locked to a specific sat, no smart contract overhead needed. elegant design honestly
jpeg_miner_ taproot sold as privacy and the first use case was JPEGs on chain. you cant script better bitcoin governance comedy
130k inscriptions in a month with zero consensus changes. say what you want about JPEGs on BTC but the engineering was elegant
the UTXO model explanation here is really well done. most articles just say ‘nft on btc’ without explaining why ordinal theory makes it possible
finally someone explains the actual inscription mechanics instead of just hypeposting. bookmarked for reference.
bookmarking this too. the UTXO explanation is genuinely one of the clearest ive read. most coverage just handwaves the technicals
utxo_nerd the UTXO model explanation was solid but the article skips how expensive inscriptions got once fees spiked. 130K inscriptions by Feb 2023 clogged the mempool and pushed fees to 100+ sats/vbyte
agreed. most ordinal coverage is either hype or FUD with zero technical substance. this is the rare piece that actually explains the mechanics
the SegWit witness discount from 4x to 1x is what made inscriptions economically viable. without that data being discounted nobody would pay 100 sats/vbyte to embed a JPEG
witness_lock_ Taproot removing the script size limit is what made this possible. SegWit alone wouldnt have been enough. the combination of witness discount plus no script cap is what created the inscription economy
embedding data directly in witness signatures instead of using a smart contract layer was genius. BTC maxis hated it but it proved the chain can do more than transfers
without Taproot this entire protocol is impossible. the upgrade quietly enabled the most significant innovation on BTC since SegWit
130K inscriptions by February and fees hit 100+ sats/vbyte by May. the technical elegance is real but the network congestion pricing out regular users was the cost nobody wanted to acknowledge
the Taproot requirement is the real bottleneck. without SegWit v1 this entire protocol would be impossible. Ordinals didnt change consensus rules, they just used what Taproot enabled in a way nobody expected
130k inscriptions in 6 weeks without a consensus change is objectively impressive engineering. whether bitcoin should host NFTs is a separate question from whether it can
130k inscriptions without a single consensus change is genuinely elegant engineering. the UTXO model doing all the heavy lifting
130k inscriptions in less than two months from launch. people who said bitcoin could not do NFTs were so confident and so wrong
taproot was the unlock nobody saw coming. that single upgrade turned bitcoin from boring store of value into a surprisingly flexible data layer
the SegWit witness discount is the real enabler here. without the 4x data discount nobody pays to embed JPEGs at 100 sats per vbyte
Anders L. exactly. Taproot removed the script size cap and SegWit discounted the witness data. without both inscriptions are economically impossible