The fee number crossed my terminal before the coffee cooled: Runes activity accounted for under 2% of total miner revenue on a rolling seven-day basis.
Twelve months ago, in the seventy-two hours after block 840,000, that same line sat near 60%.
I don't trust a single explorer, so I pulled three. Same shape on all of them. Mempool drifting around 2 sat/vB on a Tuesday afternoon. Wallets that were minting eight times a day through April have been silent for months. The Ordinals marketplace volume that used to clear nine figures a week now clears what a mid-tier NFT project did in its worst quarter.
Bitcoin's token-minting economy has stopped paying for its own blockspace, and almost nobody in the Runes camp wants to print that sentence.
That is the finding. Everything below is me trying to figure out whether it's a funeral or a detox.
Why This Number Was Always Going to Break
Context first, because the setup matters more than the collapse.
Runes launched at the halving in April 2024. The pitch was clean: a UTXO-native fungible token standard that avoided the worst of BRC-20's mess โ no endless orphaned inscriptions, no JSON sprawl jammed into witness data, less junk dumped into the UTXO set. Case and point. Technically, it was the better mousetrap.
The timing was the problem.
Runes shipped into a subsidy cut. Block rewards dropped from 6.25 to 3.125 BTC overnight, and miners suddenly needed fee revenue to hold margin. Any new demand for blockspace looked like salvation. So the market priced Runes as infrastructure. It was actually a queue.
Because what filled those blocks in the first six weeks wasn't collectors. It was farmers โ thousands of wallets racing to mint early so they could claim the airdrops that Runes issuers were handing out to early minters. The token was never the product. The snapshot was the product.
And snapshots end.
The Data: A Fee Engine Running on Fumes
Here's what I pulled from my own node plus two commercial explorers, cross-checked against public miner revenue dashboards.
Fee share of miner revenue from Runes and inscriptions combined: peaked near 61% in the first week post-halving. Now sitting in the low single digits on most days. There are still spikes โ a launch here, a mint there โ but a spike is not a baseline.
Median fee rate: from 300+ sat/vB during the April 2024 congestion down to sub-2 sat/vB. Blocks are being mined half-empty again. That is a real, measurable, on-chain fact.
UTXO set growth: this is the one I care most about. Runes was sold as the standard that would keep the UTXO set tidy. In practice, minting campaigns created millions of small, low-value outputs that were never worth the fee to spend. Dust that will sit there until fees drop low enough to sweep โ which, ironically, is now. Cheap blockspace is giving people the first real window in a year to actually clean up the mess they made.
Mint cost versus current market cap: I ran this on the top fifteen Runes tokens by mint volume. In eleven of them, the aggregate cost of minting โ fees paid, not capital deployed โ exceeded the token's present fully diluted value. Read that again. The act of creating the asset cost more than the asset is worth. That is not a market. That is a bonfire with a chart attached.
I've seen this movie. In 2017 I was filing Vietnamese-language ICO breakdowns within twenty-four hours of a whitepaper dropping, chasing the green candle through the ICO fog, and I watched dozens of projects where the gas spent minting the tokens dwarfed anything the tokens ever did. Runes didn't invent the pattern. It just moved it onto Bitcoin's ledger, where the receipts are permanent and public and slightly embarrassing.
Where the Volume Actually Went
Pulse check on the volatile heartbeat of the exchange side, because that's my day job.
When Runes liquidity was hot, it pooled in two places: a handful of dedicated Bitcoin-native marketplaces, and the wrapped-asset venues on EVM chains that listed top Runes pairs to capture the spillover. Liquidity flows where the heat is highest, and for about six weeks the heat was on Bitcoin.
Now look at the order books. Spreads on the top Runes pairs have widened to the point where a retail-size order moves price 4-6%. Market maker inventory is thin. Maker rebate programs that were subsidizing Runes pairs in mid-2024 have quietly ended or been repurposed toward the ETF-adjacent flow.
That last detail is the tell. Exchanges do not abandon a category because it's quiet. They abandon it because the customer acquisition cost stopped making sense. When a venue stops paying makers to quote your asset, they've already done the math you're about to do manually.
Miners did the same math faster. Hashrate kept climbing through the fee collapse โ the network is stronger than it has ever been while collecting a smaller fee share than at any point in the last two years. Which means the fee crash didn't hurt Bitcoin. It just removed a subsidy that miners had briefly, wrongly, counted as structural.
The Contrarian Read: Cheap Blockspace Is the Product
Here's the angle you won't find in the post-mortems.
Everyone is framing the Runes collapse as a bearish signal for Bitcoin. I think that's backwards, and I think the people framing it that way are the same people who spent 2024 telling you Bitcoin was becoming a settlement layer for digital artifacts.
It was never that. Bitcoin's core value proposition is cheap, predictable, credibly neutral settlement. Fee spikes are a tax on that proposition. A $40 transaction to move your own coins during a mint frenzy is not adoption. It is a congestion event, and congestion events are what Bitcoin's fee market is designed to eventually price out.
Sub-2 sat/vB is not a failure state. It is the product returning to its default condition after a year of being rented out.
Second blind spot: the tooling survived even though the farmers left. Ordinals indexers, Runes indexers, inscription-aware custody โ all of it is still running, still maintained, still deployed on exchange backends. That infrastructure does not evaporate when the airdrop economy does. It just sits there, waiting for a use case that isn't a snapshot.
And here's the third piece, the one I keep circling back to. The next wave of demand for Bitcoin-native digital assets is not going to arrive through a marketplace for degens. It's going to arrive through licensed venues. Watch what Hong Kong and Singapore are actually building right now โ custody frameworks, licensing regimes for digital artifact trading, compliance rails for tokens that live on a public chain. That buildout is not really about embracing innovation. It's about who ends up holding the assets when the institutional money finally wants somewhere regulated to put them. Amidst the noise, the smart money whispers, and right now it's whispering in Cantonese and English, in rooms with compliance officers.
Whoever wins that licensing race captures the next cycle's flow. The Runes chart is a distraction from that fight.
What I'm Watching Next
One signal matters more than the rest: the next difficulty adjustment, and the ratio of inscription fees to total miner revenue in the seven days after it.
If that ratio stays in the low single digits while hashrate holds, then the fee market has genuinely reset and the Ordinals experiment has entered its quiet, functional phase โ the one where nobody writes about it and somebody quietly builds on it.
If it spikes again on another airdrop, you'll know the cycle is still running on fumes.
From frenzy to function: tracing the cycle is the only honest work left in this market. The question isn't whether Runes dies. It's whether the tooling outlives the hype โ and whether we notice before the next launch.