Omni

Omni

The Mastercoin-lineage ledger, TetherUS included. BTI decodes Class C messages in OP_RETURN itself; a ledger token's bytes are its operations.

Omni 0.86 % of the chain's bytes in Book A, as of block 972,299, 2026-10-20 09:30:48 UTC

An Omni property is a token of the Omni protocol, of the Mastercoin lineage. BTI's token table names it by its property id, assigned by replaying the protocol. A human name is an alias: TetherUS is property 31, and BTI writes it as omni/31.

How Omni puts data on Bitcoin #

The data-types reference lists three Omni channels over time:

Class A (named in the reference as Mastercoin Class A): data in fake pay-to-pubkey-hash addresses, 20 bytes per output, with 546 sats of dust each. The reference calls the general trick undetectable by construction.
Class B: data in fake public keys inside bare multisig outputs, which stay in the UTXO set until spent.
Class C: a message in an OP_RETURN output.

Case 17 of the BTI case studies decodes a Class C simple send at block 560,000. The OP_RETURN output is 31 bytes, its script 22 bytes, beginning with the marker 6f6d6e69 ("omni"). The payload reads version 0, type 0 (simple send), property 31 (TetherUS), and an amount of 738,052,430,000 base units. A second, 546-sat output is the reference output: the recipient.

How BTI attributes the bytes #

Omni is a native protocol: BTI's own decoder reads the chain, and there is no external indexer. The decoder reads Class C only. The adapters table says that is 99.67 % of Omni's bytes. Ruleset v1 names Class B as a coverage gap (B4b). omni-core is on the watchlist; deploying it would move Omni from native toward reconciled.

Omni carries no file. The books document gives it as the pure-ledger shape: "content 0 honestly; structure and overhead carry the backing — a ledger token's bytes are its operations".

The case studies show both sides of the dedication test:

Dedicated (case 17, block 560,000). Every output serves the send or is change, so the claim covers the whole 257-byte transaction: content 0, structure 257, overhead 0, and the whole 50,000-sat fee. The case reconciled with the live database to the byte.
All riding (case 21, block 550,000). A TetherUS send in a transaction that also paid 48,625 sats that no Omni rule required. The dedicated test answers no, so the claim keeps only its frame, the reference output and its carrier: 75 bytes. With no dedicated claimant, the other 328 bytes "stay with the transaction", bought by nobody. In the ruleset's words, "every claimant aboard was a passenger".

Live figures #

The family's books (as of block 972,299, 2026-10-20 09:30:48 UTC):

measure	value
Book A	6.66 GB
share of the chain's bytes (Book A)	0.86 %
share of all Book A bytes	4.98 %
properties	404
operations	19,432,030
first height	not measured
last height	971614

The protocols API gives no first height for Omni, so this page says not measured rather than zero.

On 2026-10-18 the heaviest tokens board placed omni/31 third of all tokens by Book A, at 6.5 GB. It showed unnamed: the board had no label for it.

Holders #

The leaderboard cannot place Omni's bytes on addresses. Read on 2026-10-20, its coverage table listed Omni as not measured, because Blockspace Holdings keeps no Omni ledger.

Measured findings #

In the July 2026 measurement run, Omni had a bytes:vbytes ratio of 1.0142, "pre-segwit, as expected": almost no witness discount. Inscriptions measured 3.374 and runestones 1.0000.
The same run lists Omni, with Counterparty and early OP_RETURN, as what was on the chain at height 600,000, when data payload was 9.2 % of block bytes.

Sources #

Live figures: protocols as of block 972,299, 2026-10-20 09:30:48 UTC. Refreshed hourly.