CKB is capacity first, money second

Nervos CKB’s core move is blunt and structural: the token is the state budget. Capacity is not an abstract “gas” unit. Capacity is literally the right to occupy bytes in the L1 state. The official capacity rules spell it out: “1 CKByte grants the right to store 1 byte on-chain”, and the sum of your unlocked cells’ capacity is your byte-budget.

That design forces tokenomics to do infrastructure work. If you want to keep a contract, an asset, or any persistent state on CKB, you tie up CKB as “occupied capacity”. If you later delete that state by consuming the cell, you can free the same CKB back into circulation.

It also makes “dust” non-trivial. A cell must reserve a minimum of 61 CKBytes just to store its essential fields, because those fields themselves take 61 bytes, and capacity is byte-denominated.

From a decentralization purist lens, this is the right axis to optimize. Keeping the state bounded and paid for reduces long-run pressure toward datacenter-only full nodes. Nervos makes token demand map to a scarce resource that every validator node must carry: state. That’s a more honest economic primitive than narratives about “value accrual”. The trade-off is that you must accept an ongoing, protocol-level “state rent” mechanism. Nervos chose targeted inflation for that job.

Supply: three sources, one hard cap, one tail

Nervos CKB has three supply components:

1) Genesis issuance. The genesis issuance block issued 33,600,000,000 CKB. Of that, 25% or 8,400,000,000 CKB was “effectively burned” at genesis and “will never enter circulation”, leaving 25,200,000,000 CKB circulating from genesis allocations.

If you want a quick refresher on issuance terminology, see our tokenomics FAQ.

2) Base (primary) issuance. Base issuance has a finite total of 33,600,000,000 CKB and halves about every 4 years, Bitcoin-style.

3) Secondary issuance. Secondary issuance is a fixed-rate tail emission of 1.344 billion CKB per year. It is not hard-capped.

CKB mainnet (“Lina”) launched on November 15, 2019.

The first base-issuance halving is documented as occurring on November 19, 2023.

For current market-facing supply stats, CoinGecko (accessed on March 5, 2026) reports a circulating supply of 48,179,894,049 CKB and a total supply of 48,942,329,709 CKB.

Genesis allocations and unlocks

Nervos publishes the genesis distribution with percentages, amounts, and unlock dates. These are allocations of the 33.6B genesis issuance, not the later base issuance mined to miners.

That “burn” is not just moral theater. Nervos explicitly treats it as a live part of the economic machine. The documentation describes the burned amount as affecting secondary issuance routing by hard-coding part of it as “occupied” and part as “liquid”. Specifically, from the burned 25%: 15% of genesis issuance (5,040,000,000 CKB) was hard-coded as “occupied capacity” and 10% (3,360,000,000 CKB) as “liquid”. This guarantees baseline secondary-issuance flows to miners and the treasury in extreme scenarios.

Secondary issuance, “state rent”, and the NervosDAO hedge

Nervos frames secondary issuance as the mechanism that charges for long-lived state without requiring explicit recurring payments. The targeted inflation scheme is described as: state occupiers pay an inflation tax because they do not receive secondary issuance, while non-occupiers can lock in the NervosDAO to get compensated and neutralize that dilution.

Mechanically, secondary issuance is split across three buckets based on how existing CKB is used: (1) “occupied” state, (2) CKB deposited in NervosDAO, and (3) liquid CKB that is neither occupied nor in the DAO. The first bucket drives miner allocation. The second bucket drives depositor compensation. The third bucket is routed to a treasury destination.

The treasury portion is the most politically sensitive flow. RFC0015 states that the “rest of the secondary issuance” is determined by the community through governance, and until agreement, it is burned.

Nervos reiterates in its distribution write-up that the treasury is not activated at launch and requires approval through governance and implementation through a hard fork, and that before activation this portion is burned.

The NervosDAO itself is a system-level smart contract that functions as a time deposit. Deposits have a minimum deposit period measured in blocks and withdrawals are a two-phase process, with the second phase constrained by a required lock period.

The decentralization angle here is not “APY”. It’s governance minimization. If long-term holders can self-serve an inflation hedge on-chain, you remove a whole class of discretionary interventions that usually become foundation-controlled “staking programs”. Nervos gets points for that. Still, it shifts power toward whichever actors can most efficiently keep capital liquid while still capturing the DAO hedge, which is an ecosystem-level distribution question, not a protocol guarantee.

Fees and miner revenue: NC-Max makes it more complex than “block subsidy + fees”

CKB fees are paid by leaving total input capacity larger than total output capacity, and the difference is paid to the miner that creates the block containing the transaction.

In the official mining docs, miner revenue is decomposed into four components: base reward, secondary reward, proposal reward, and commitment reward.

Base reward is the halving subsidy stream. The docs restate the headline constants: 33.6B CKB total base reward, halving every 4 years, with the first halving on November 19, 2023.

Secondary reward is the tail emission stream: 1.344B CKB annually, distributed based on network usage rather than purely to miners.

Then there is the part most analysts skip because it is operationally annoying. Transaction fees are split between proposal and commitment rewards under NC-Max. In the docs: miners of block N receive 40% of fees for transactions proposed in block N and committed from N+2 to N+10, and 60% of fees for transactions proposed earlier and committed in block N. All rewards are allocated at N+11.

Two decentralization implications fall out of this.

First, miner economics are not allowed to collapse into fee-only security. That is the explicit purpose of secondary issuance, and it is coherent for a preservation-first L1. The cost is permanent protocol inflation for actors who refuse or cannot use the DAO hedge.

Second, real-world validator distribution is mostly a mining distribution question, not a staking distribution question. CKB uses Proof of Work, and its mining guide states it uses the Eaglesong hashing algorithm and that mining “today requires specialized ASIC miners”.

Permissionless entry exists at the protocol level. Industrial concentration still happens at the hardware and pool layers. That’s not a Nervos-specific sin. It is the standard PoW trade-off: you buy credible neutrality at the consensus layer and you pay for it with supply-chain and scale pressures in the mining layer.

Governance surfaces and where decentralization can actually fail

Nervos’ monetary policy is presented as fixed. The issuance-model documentation describes the monetary policy and issuance schedule as “completely fixed” and “not subject to any change”, framed as an invariant of the network.

As a decentralization purist, I treat that claim as aspirational, not binding. In any open-source chain, parameters are fixed until enough economic weight coordinates a fork. What matters is the coordination topology.

Nervos’ docs describe a three-phase hard fork process, and explicitly treat RFC finalization as part of the path to activation. It also states that if most nodes upgrade by the specified epoch number, the hard fork is successfully deployed, and old nodes will no longer connect.

That is normal for PoW governance. It is also where decentralization risk concentrates, because “most nodes” is not a threshold with a formal constituency. It is a coordination outcome.

There are two governance choke points that matter for tokenomics:

1) Treasury activation and calibration. RFC0015 defines the “rest of the secondary issuance” as governed by the community, and burned until agreement.

The genesis distribution write-up also states the treasury will be activated only with governance approval and implemented through a hard fork.

Until activation, the protocol is effectively self-denying an on-chain discretionary budget. That reduces governance surface area. It also means ecosystem funding leans harder on the genesis allocations and whatever off-chain coordination can assemble around them. That is a decentralization trade-off, not a free lunch.

2) Protocol upgrades that touch economics. Hard forks are epoch-triggered upgrades that require broad upgrade coordination.

In practice, “governance” here is the intersection of miners, exchanges, infrastructure operators, and large holders, plus the credibility of the RFC process. That is workable. It is not formally decentralized in the way on-chain voting systems pretend to be. It is social consensus with a strong infrastructural veto.

For a tokenholder-voting contrast, see our ENS tokenomics.

Risk register (ranked)

Dominant risk: mining centralization becomes de facto governance, and the tokenomics amplify the feedback loop.

CKB is PoW. That is the right base layer choice if your north star is censorship resistance and minimizing plutocratic “stake governance”. But PoW decentralization is not a vibe. It is a distribution problem with a moving target: ASIC supply, energy cost geography, pool coordination, and latency advantages. Nervos’ own mining guide explicitly states that mining “today requires specialized ASIC miners”. That is already a centralizing force versus CPU-minable designs, even if it is rational for security.

Here is the tokenomics-specific kicker. Nervos adds secondary issuance to fund miners in perpetuity, and it explicitly ties the miner share of secondary issuance to state occupancy. The mining docs frame secondary issuance as a fixed annual stream and reiterate it is distributed based on network usage, not handed entirely to miners by default.

That sounds balanced. The failure mode is subtle and structural.

If a small set of pools dominates block production, they are not just earning most base issuance. They are also controlling the capture of the secondary issuance component that is sensitive to how “occupied vs DAO vs liquid” is measured at any point in time. That measurement is mechanical, but the ecosystem behaviors that drive it are not. Wallet UX, exchange custody policies, and large-holder coordination determine how much supply sits in DAO, how much is left liquid, and how much is tied up in state. Those are soft levers that can be influenced by concentrated actors without changing a single protocol constant.

Now add governance. The “rest of secondary issuance” is routed to a treasury destination that is currently burned until governance decides otherwise. RFC0015 explicitly says governance determines its use, and it is burned until agreement.

Activating that treasury requires a hard fork, and hard forks are deployed when “most nodes upgrade by the specified epoch number”. That is an infrastructure coordination threshold, not a tokenholder vote threshold.

So the dominant risk is not “inflation”. It is governance capture through validator concentration. Once mining is meaningfully concentrated, the same actors that decide what chain is “canonical” in practice also become the strongest veto players over treasury activation, treasury policy, and any economic hard fork. That is how decentralization fails on PoW networks. Quietly. Through operational coordination and economic dependence.

Indicators you can actually watch are simple. If hashrate concentrates into a few pools, and if exchange-custodied balances become the dominant DAO participants, the real governance constituency shrinks. If, at the same time, treasury activation becomes politically necessary because genesis allocations are depleted, the system is forced into a high-stakes coordination moment where concentrated infrastructure wins by default.

None of this is guaranteed to happen. The point is that Nervos’ tokenomics make miners structurally important forever. That is intended. It also means the network cannot afford miner centralization, because miners are not a temporary bootstrap actor. They are the permanent security budget recipient.

If you like PoW, you should want this risk named plainly. The mitigation is not marketing. It is continuous social pressure toward pool diversity, credible neutral client implementations, and UX that makes DAO usage broadly accessible so dilution protection is not an insiders-only feature. The protocol can help. The ecosystem has to execute.

  1. Mining centralization drives governance capture. Trigger: sustained concentration of block production into a small number of pools or operators. Mechanism: PoW “who mines blocks” becomes “who decides upgrades”, especially for treasury activation via hard fork. Who bears it: users, app teams, and minority miners who lose credible neutrality and censorship resistance. Measurable indicators: rising top-pool hashrate share, declining number of meaningful pools, and a growing share of issuance routed to a small set of payout addresses.
  2. DAO hedge becomes socially inaccessible, creating a two-tier holder class. Trigger: custody defaults (exchanges, wallets) that discourage or block NervosDAO usage. Mechanism: non-DAO holders are diluted by secondary issuance while sophisticated holders neutralize it, shifting long-run distribution toward insiders. Who bears it: retail holders and passive participants. Measurable indicators: DAO participation rates stagnate while circulating supply grows, plus increasing concentration of DAO deposits in a small number of addresses.
  3. Treasury activation becomes a legitimacy crisis. Trigger: ecosystem funding needs rise after genesis ecosystem allocations are spent, increasing pressure to redirect the “burned until agreement” stream. Mechanism: a contentious hard fork or informal deal-making among concentrated actors determines treasury policy. Who bears it: the whole network through coordination risk, chain splits, or perceived capture. Measurable indicators: public RFC and forum contention on treasury parameters, rising reliance on foundation-led grants, and repeated calls to activate treasury via hard fork.

If you need help stress-testing these dynamics for a new app that will occupy meaningful state, this is where token economy design stops being academic. A short tokenomics consulting engagement can quantify how much capacity you truly need, how quickly you can recycle it, and what your expected secondary-issuance exposure looks like under different DAO behaviors. Keep it mechanistic. The chain already is.



This article is part of our Tokenomics Deep Dive series.