BRC-20 and the metaprotocols

Wonder

1. What happens when a token standard relies on third-party servers to track balances instead of Bitcoin’s own consensus?
2. How did a rush to mint small tokens clog the mempool in the spring of 2023?
3. Why does the creation of millions of tiny outputs pose a specific threat to the long-term health of the UTXO set?

Explore

Part IBRC-20 tokens

1. Bitcoin’s protocol ignores the meaning of data embedded in transactions; it checks only that spends are valid. Inscriptions (the previous lesson) showed that anyone could write arbitrary content into Taproot witness data without changing consensus rules. BRC-20 took the next step: using that same channel to run a fungible token system that bitcoin itself knows nothing about.
2. BRC-20 emerged in March 2023 as an experimental token standard built directly upon the Ordinals inscription mechanism. Created by the pseudonymous developer Domo, it uses simple JSON objects to manage fungible tokens. A ‘deploy’ inscription defines a new token with a maximum supply and a limit per mint; subsequent ‘mint’ inscriptions increase individual balances, while ‘transfer’ inscriptions move them between wallets. Because Bitcoin nodes do not validate this data for consensus purposes, BRC-20 functions as a sidecar application that exists only in the interpretation layer.
3. The critical architectural difference from native assets is that Bitcoin itself does not track BRC-20 balances. Off-chain indexers must scan the entire blockchain history to catalog every deploy, mint, and transfer inscription to compute a user’s holdings. This separation means a ‘balance’ is a calculation performed by an external service rather than a fact verified by Bitcoin’s proof-of-work consensus. Different indexers may disagree on state if they process transactions differently or experience downtime, introducing a layer of dependency that native bitcoin does not have.
4. In the spring of 2023, speculative interest in BRC-20 tokens led to a historic surge in transaction fees on the Bitcoin network. Thousands of users attempted to execute ‘mint’ transactions simultaneously to claim their allocation before the global supply cap was reached. This competition for block space pushed fee rates (in satoshis per virtual byte) to their highest average in two years, and total fees in some blocks exceeded the block subsidy for the first time since 2017, clogging the mempool and delaying standard payments. The event demonstrated how an external application layer could significantly impact the economic dynamics of the base layer.
5. Critics argue that BRC-20 is a highly inefficient use of Bitcoin’s limited block space compared to other asset protocols. By storing simple JSON strings in witness data, it consumes capacity otherwise available for standard cash transfers, effectively taxing all network users through higher fees. Furthermore, the many small outputs created by frequent minting and transferring actions added substantial bloat to the Unspent Transaction Output (UTXO) set, increasing storage requirements for full nodes. Supporters counter that this activity represents legitimate demand for bitcoin block space, providing new revenue streams for miners as the subsidy halving approaches.
6. Open a block explorer and inspect a BRC-20 mint transaction from May 2023. The raw JSON data is embedded in the witness of the input spending the Taproot output. Querying a standard Bitcoin wallet for ‘BRC-20 balance’ returns zero unless connected to an external indexer like Unisat or Magic Eden, highlighting that the token exists only in the interpretation layer. This disconnect between on-chain data and off-chain accounting is the defining characteristic of the BRC-20 experiment.

Part IITokens on Bitcoin: Ordinals, BRC-20, Runes and programmable metaprotocols

7. Casey Rodarmor launched Ordinals in January 2023, assigning a serial number to every satoshi based on its position in the chain. This allowed data to be inscribed onto specific satoshis using Taproot script paths. The protocol treated each unit as unique, similar to non-fungible tokens on other chains. Bitcoin nodes recorded the transaction but did not validate the inscription’s content or enforce any token logic beyond standard UTXO rules.
8. In March 2023, Domo introduced BRC-20, using JSON inscriptions within Ordinals to create fungible tokens via deploy, mint, and transfer functions. This approach was clumsy; it relied entirely on off-chain indexers parsing the JSON to maintain balances. Different indexers could diverge on state if they missed transactions or interpreted rules differently. The standard consumed significant block space without adding consensus-level security for the token itself.
9. Rodarmor launched Runes at block 840,000 on April 20, 2024, coinciding with the halving to 3.125 BTC. Runes transfers are described by protocol messages called runestones, in OP_RETURN outputs; the balances themselves sit on ordinary UTXOs. Unlike BRC-20’s JSON, Runes use a compact binary format parsed by indexers. A transaction may hold balances of multiple runes in its outputs. The protocol is UTXO-based, meaning rune balances are attached to specific outputs and move from inputs to outputs.
10. All these protocols are metaprotocols: Bitcoin records the data on-chain, but off-chain software interprets the rules. Consensus nodes do not track token balances or enforce mint limits; indexers do. This creates a trust model where the settlement layer (Bitcoin) is secure, but the application logic depends on indexer agreement. If an indexer fails or acts maliciously, users may lose access to their perceived balance even though the underlying satoshis remain intact.
11. Newer metaprotocols like Alkanes extend this by executing WebAssembly contracts within indexers, adding programmability beyond simple token transfers. This increases complexity and deepens reliance on off-chain computation. The trade-off remains constant: users gain expressive features at the cost of moving rule enforcement away from Bitcoin’s consensus layer. Settlement is final at the Bitcoin level, but state consistency depends entirely on the indexer ecosystem.
12. To verify this, open a block explorer that supports Runes and inspect any transaction containing an OP_RETURN output with runestone data. The runestone encodes rune IDs (in BLOCK:TX format) and amounts. Then check a standard Bitcoin node; it will validate the signature and UTXO structure but display no token balance information, confirming that the token state exists only in the indexer’s database.

Key facts

Exercise nine, six, three, one

Fact check

Dates, figures and claims checked against the published sources listed below.

• Domo, 'BRC-20' documentation (March 2023)
• Bitcoin Optech newsletter coverage (2023)
• mempool.space fee history data (Spring 2023)
• C. Rodarmor, Ordinals documentation (docs.ordinals.com, 2023)
• domo, BRC-20 experiment documentation (March 2023)
• C. Rodarmor, Runes documentation (April 2024)
• A. M. Antonopoulos & D. A. Harding, Mastering Bitcoin, 3rd ed. (2023)
• Alkanes protocol documentation