Blockchain knowledge firm Bitquery revealed a census on Sept. 2 of 4 methods textual content seems in Bitcoin, scanning 965,135 blocks and revealing that the identical phrases can impose very totally different prices on nodes.
Its research counted 13,062 OP_RETURN taunt transactions that added no spendable state, towards 96,231 fake-address textual content outputs that stay within the unspent transaction output set.
The March 21-26, 2026, marketing campaign named Bitcoin Core developer Luke Dashjr in the identical two-sentence taunt. Bitquery counted 13,062 transactions from 13,062 sending addresses throughout 579 blocks, with 0.1014 BTC paid in mixture charges. The chain information the textual content, timing, inputs, and charges, but it surely doesn’t establish the payer, creator, or motive.
Throughout the broader scan, Bitquery categorised 4,412,782 OP_RETURN outputs as readable textual content. Their node price differs from fake-address textual content as a result of Bitcoin treats the 2 output sorts otherwise after affirmation.
Nodes don’t must retain an OP_RETURN output within the UTXO database used to test whether or not cash might be spent, though its knowledge stays in Bitcoin’s block historical past. The Bitcoin developer information describes this null-data development as preferable to strategies that create outputs which look spendable and burden the UTXO set.
Pretend-address textual content locations readable bytes the place an handle hash would usually go. A node sees an apparently spendable output, whereas the embedded textual content has no recognized corresponding non-public key.
Bitquery categorised 96,231 such outputs throughout 3,286 transactions and estimated that they maintain about 3.2 BTC that’s successfully unspendable. These outputs persist as present node state until they’re spent.
Bitcoin Core 30.0, launched in October 2025, made bigger OP_RETURN transactions simpler to relay underneath its default coverage. The discharge notes raised the default -datacarriersize setting to 100,000 and allowed a number of OP_RETURN outputs underneath an mixture restrict for relay and mining.
These standardness decisions didn’t change the consensus guidelines that decide whether or not a block is legitimate.
Bitquery discovered that messages longer than the previous 80-byte payload ceiling appeared only some dozen occasions in most months earlier than v30 and rose into the hundreds by February 2026. Simpler default propagation affords one mechanism for that rise, however the observational sequence doesn’t present that v30 alone prompted it.
The March taunt disclosed textual content is 62 ASCII bytes, which already match beneath the previous ceiling. The marketing campaign measures the low price of repeating OP_RETURN textual content after v30.
Bitquery additionally lacked coinbase fields for 191,427 historic blocks and didn’t rely witness inscriptions. The research subsequently compares the state penalties of 4 measured channels relatively than tallying every bit of textual content embedded in Bitcoin.
Inside that scope, OP_RETURN expands block historical past with out bloating the UTXO set, whereas fake-address textual content creates the lasting node-state burden.
