Midnight makes fees a renewable resource, not a token you spend

Midnight’s core economic move is structural. Transactions are paid in DUST, a shielded, non-transferable network resource, while NIGHT stays as the transferable capital and governance asset. Holding NIGHT continuously generates DUST. You do not spend NIGHT to transact, per the NIGHT token details.

That split changes almost everything that normally dominates token design. First, it decouples “using the chain” from “selling the chain’s token.” Second, it replaces fee markets that distribute value to validators with a resource metering system where DUST is burned on use. Third, it forces the security budget discussion into a single place: Reserve-driven block rewards, not user-paid fees, as specified in the tokenomics whitepaper.

If you like deterministic mechanism design, Midnight’s best idea is also its biggest constraint-and it’s a useful case study in token economy design. The strongest parts are the protocol-enforced accounting invariants and closed-form reward formulas. The weakest parts sit where the docs still lean on “configurable parameters” and future governance.

What NIGHT is, and what it actually does

NIGHT is the unshielded native token of the Midnight network and is also positioned as the governance token. For a contrasting governance-token profile, compare HTX DAO tokenomics. Its primary functional utility is to generate DUST, which is then used to pay for transaction execution and smart contract operations.

The whitepaper is explicit about the direction of utility. DUST is the thing you spend. NIGHT is designed to be non-expendable for transactions and to remain intact as a long-lived asset whose balance drives DUST availability over time.

Two details matter for implementation realism.

One, NIGHT exists natively on both Cardano and Midnight. It is minted on Cardano as a Cardano Native Asset, then mirrored on Midnight at genesis.

Two, the system is designed so NIGHT cannot be “unlocked on both chains at once.” The whitepaper models this with explicit cross-chain state accounting and an invariant that bounds unlocked supply across chains (including an inequality form that “errs on the side of fewer tokens unlocked” to prevent double-spend exploits).

Subunits are defined in the whitepaper. 1 NIGHT = 1,000,000 STAR.

Supply, pools, and distribution: fixed cap, multi-phase community-first, and a thaw

Total supply is fixed at 24,000,000,000 NIGHT, minted on Cardano and mirrored onto Midnight at genesis.

NIGHT’s token launch on Cardano is dated in the launch redemption guide. On December 4, 2025, NIGHT was launched as a Cardano Native Asset and moved into the redemption flow for community allocations.

The distribution is organized as a “Glacier Drop” with multiple phases, including a claim phase targeting self-custody holders across multiple ecosystems, a public Scavenger Mine phase, and a Lost-and-Found phase that begins at Midnight genesis and remains available for years.

Key dates and eligibility are published by Midnight. The Glacier Drop snapshot date was June 11, 2025, and the initial claim window ran August 5, 2025 through October 20, 2025, for self-custody holders of ADA, BTC, ETH, SOL, XRP, BNB, AVAX, and BAT.

Midnight reports over 3.5 billion NIGHT claimed by individuals in the Glacier Drop claim phase, and 1 billion NIGHT claimed in the Scavenger Mine phase with over 8 million participating addresses.

The Scavenger Mine distribution also has an explicit, deterministic allocation rule at the individual level. Per the whitepaper, a participant’s daily reward is proportional to their share of completed computation units relative to the day’s total (Rm = Ar × Cm / C).

Thawing and redemption mechanics are also concrete. Community-claimed tokens unlock in four equal installments (25% each), with a randomized first unlock day in the initial 90-day window, then subsequent unlocks every 90 days.

There are two durations that people often conflate. The whitepaper defines a 450-day Redemption period, which includes 360 days of thawing plus a 90-day grace period.

In launch communications, Midnight sets the operational dates. Thawing for community allocations begins December 10, 2025 and ends December 4, 2026, followed by a 90-day grace period.

The project also published a “final launch allocation” breakdown by pool. These amounts sum to the 24 billion fixed supply.

The deterministic takeaway is that Midnight’s supply policy is simple at the top level. Fixed cap. No fee-minting. Emissions are a transfer from Reserve into circulation via block rewards until the Reserve depletes, at which point circulating supply matches total supply and the protocol rules are meant to prevent any further creation.

DUST mechanics and fee flow: burn-on-use and dynamic congestion pricing

DUST is not a second token in the usual sense. Midnight describes it as a shielded network resource used to pay for transaction execution. It is non-transferable, renewable while designated by NIGHT, and decays when the association is severed or the associated NIGHT balance drops.

DUST generation is address-based and intentionally separable from the NIGHT address. A NIGHT holder designates a DUST address as the recipient, after which DUST accumulates linearly per block until it hits a cap. The DUST recipient address is DUST-specific and non-derivable from the originating NIGHT address.

The cap is proportional to the associated NIGHT balance. When the DUST cap is reached, generation halts until DUST is used, after which it resumes back up to the cap. When DUST is used for a transaction, it is expended and burned. Midnight also states that transaction fees are not collected by block producers and there is no DUST tipping mechanism.

Transaction pricing is dynamic. The whitepaper defines fees as composed of three parts: a minimum fee, a congestion rate, and a transaction weight.

The core fee equation is given directly as TxFeen = CongestionRaten × TxWeight + MinFee.

The congestion rate is updated multiplicatively from its previous value using a “fee adjustment factor,” and is described as a DUST-per-byte multiplier applied to the transaction’s weight.

Midnight sets a 50% block utilization target as a system parameter that can be adjusted via governance action. This target is directly tied to the dynamic pricing logic, since fees rise when blocks exceed the target and fall when utilization is below it.

As a mechanism designer, I like the clarity of the fee surface. Weight times congestion plus a floor. The trade-off is equally clear. If the chain succeeds, DUST becomes scarce in the exact way block space is scarce. That is the design. The question is whether the “cost predictability” narrative holds in sustained high demand, given that the congestion term is explicitly designed to move.

Security budget and emissions: Reserve-funded block rewards with a closed-form curve

Midnight’s validator incentive system is not fed by user fees. DUST fees are burned, and the reward stream comes from a protocol-managed Reserve funded from the fixed NIGHT supply. Block rewards are dispensed as a fixed percentage of remaining Reserve tokens per block, which creates a smooth, tapering curve as the Reserve decreases.

The reward math is specified with explicit variables. The base reward per block is Nb = Bo × R, where Bo is the current outstanding token balance in the Reserve and R is the base distribution rate.

The whitepaper defines the base distribution rate in terms of an annualized expansion of circulating supply, then divides by expected blocks per year. It gives Ra = π(1 − B − T) and R = Ra ÷ γ, where B and T are the Reserve and Treasury allocations as fractions of total supply, π is the annual inflation rate for circulating supply expansion, and γ is the number of blocks per year.

Rewards then split between block producers and the Treasury based on two parameters.

The first is the subsidy rate S, which determines the fixed portion of the base reward that goes to the block producer regardless of utilization. The fixed component is Nf = Nb × S.

The second is block utilization U, which determines how much of the variable component goes to the producer versus the Treasury. The variable component is given as Nv = U × (Nb − Nf), and the producer’s actual reward is Na = Nf + Nv. Whatever remains, Nt = Nb − Na, is sent to the Treasury.

The whitepaper sets subsidy rate S = 95% at launch, and states it can be adjusted via governance action, with an expectation that it could move toward 50% in the future.

That 95% launch setting is a blunt, rational choice. It bootstraps the supply-side of security before demand exists, and it reduces the incentive to stuff blocks with valueless transactions to capture variable rewards. The cost is that Treasury inflows from underfilled blocks are structurally small at first. You should treat “Treasury-backed token” narratives as forward-looking until the Treasury is both unlocked and meaningfully funded beyond its initial allocation.

There is also a sequencing dependency. Block production at launch is described as permissioned, with a progressive path toward permissionlessness. The initial permissioned block producers do not receive rewards. Over time, Cardano SPOs can become Midnight block producers and become eligible for NIGHT rewards, with selection proportional to delegated ADA stake and registration via a smart contract on Cardano.

Governance and parameter control: multisig first, “full spec later”

The whitepaper is direct that Midnight will start with a federated governance structure. A select committee of stakeholders with equal governance powers can submit and vote on proposals and protocol upgrades. Passage requires a threshold of approvals, implemented via a multisig mechanism.

It also states that the full specification and mechanics of decentralized on-chain governance are expected to be detailed in a future document.

Two immediate implications follow.

First, several economically critical parameters are explicitly adjustable “via governance action.” That includes the block utilization target that shapes DUST fee dynamics, and the subsidy rate that shapes how rewards split between block producers and the Treasury.

Second, some of the most value-sensitive token functions are not available at mainnet launch. The whitepaper states that neither using NIGHT for governance purposes nor using NIGHT from the on-chain Treasury for ecosystem growth initiatives will be available at mainnet launch, and these are expected to be implemented in the future.

From a deterministic-rules perspective, this is the main tension in the design. Midnight has unusually explicit formulas where it matters for day-to-day operation. Yet the parameter surface is wide, and the early control plane is a committee multisig with committee composition “yet to be identified or formed.” That is governance adaptability. It is also parameter instability risk, because constraints are not fully specified today.

We track similar control-plane risks in our research reports.

Risk analysis

Midnight’s token system is mechanically coherent, but it is still early in the parts that usually harden last: decentralized governance, Treasury activation, and the transition from a staged launch into a stable long-run policy regime. The project’s strongest move is also its biggest modeling trap. DUST makes transactional costs legible in protocol units, but it does not eliminate congestion dynamics. It relocates them.

Top 3 risks

  1. Governance-driven parameter drift (dominant risk). Trigger: a federated committee changes key economic parameters before decentralized governance is specified and live. Mechanism: multiple critical “system parameters” are explicitly adjustable via governance action, including DUST fee dynamics (block utilization target) and reward splitting (subsidy rate), and the initial governance is a multisig committee with composition and thresholds not fully specified in the tokenomics document. Who bears it: NIGHT holders (policy credibility), builders (fee predictability), and block producers (reward predictability). Measurable indicators: published governance actions that change the block utilization target or subsidy rate, changes to protocol core parameters like block size, and delayed publication of a full on-chain governance specification.
  2. Utility-to-demand mismatch for NIGHT. Trigger: DUST access becomes abstracted via sponsorship and capacity market intermediaries, while direct holding demand for NIGHT does not rise proportionally. Mechanism: the protocol explicitly supports models where end users transact without holding NIGHT, including sponsorship and “tokenless” abstractions, and it contemplates Babel Stations that submit transactions on users’ behalf in exchange for non-NIGHT payment. Who bears it: NIGHT holders (weaker direct demand linkage), and Treasury policy designers (harder to align value capture). Measurable indicators: share of transactions sponsored by third parties, growth of DUST leasing or Babel Station usage, and the ratio of active DUST-designating NIGHT addresses to total active accounts.
  3. Distribution overhang and redemption complexity. Trigger: large-scale unlock events coincide with thin organic demand, or redemption UX breaks for long-tail participants. Mechanism: community allocations thaw in four 25% tranches over 360 days with randomized start dates, and redemption requires repeated interactions during the active period, plus a grace period before the portal sunsets. Who bears it: claimants (operational risk), liquidity venues (flow volatility), and long-term holders (price impact from distributed unlocks). Measurable indicators: redemption participation rates per tranche, token flows around tranche boundaries, and the volume of unredeemed allocations approaching the end of thawing and grace windows.

Dominant risk: governance-driven parameter drift

Midnight’s whitepaper does two things at once. It gives you closed-form formulas for the security budget and the fee engine, and it explicitly flags many of their inputs as configurable. That is not an accident. It is a recognition that early networks need a control surface to respond to abuse patterns and adoption reality.

The problem is not that parameters can change. The problem is that the system currently lacks a fully specified, credibly constrained process for changing them. The whitepaper says the initial governance will be federated, run by a committee of equal powers, gated by a multisig threshold. It also says key entities are “yet to be identified or formed,” and that the full on-chain governance spec will come later.

In mechanism terms, this is a temporary discretionary regime sitting on top of deterministic mechanics. That mismatch matters because Midnight’s value proposition is predictability. It sells a fee model that is more legible than gas, and an emissions curve you can compute. But if the committee can change the block utilization target that drives dynamic pricing, the “predictability” claim becomes conditional. It becomes “predictable given current committee preferences.”

There is a second-order effect. Midnight’s Treasury is designed to accumulate NIGHT (initially) and then potentially diversify via capacity marketplace fees. Yet Treasury usage is explicitly unavailable at mainnet launch, and Treasury tokens are initially locked. That means a large portion of the “collective decision-making” and “collective capital allocation” story is structurally deferred. During the deferral window, the committee’s choices effectively substitute for tokenholder governance, while tokenholder governance is not yet active.

Finally, governance risk is not only “can they change monetary policy.” The whitepaper lists protocol upgrades, hard forks, and core parameter updates as committee responsibilities. Those can change fee pressure, state growth constraints, and even the practical competitiveness of node operation. A chain can keep the same formulas while changing the environment in which those formulas produce outcomes.

What would reduce this risk in a way a mechanism engineer respects is not “good intentions.” It is constraints. Explicit bounds on parameter changes per epoch. Timelocks. Formalized emergency powers with expiry. Public, testable rules for when a parameter can move and by how much. The current docs do not specify these constraints. That leaves you with structural uncertainty about parameter stability even if the underlying formulas are clean.

If you are doing tokenomics consulting or token economy design work around Midnight-like dual-asset systems, you can model the equations quickly. The hard part is governance realism. Price in the fact that the control plane, not the math, is the dominant source of variance until decentralized governance is specified and deployed.



This article is part of our Tokenomics Deep Dive series.