STRK stopped being “governance optional” and became operationally mandatory

Starknet’s token design made a clear pivot when transaction fees became STRK-only with the release of v0.14.0 on September 1, 2025, as described in the STRK token docs. That single change matters more than any airdrop headline because it hard-wires ongoing demand for STRK into the chain’s basic function.

In the project’s own framing, STRK is meant to cover three pillars: fees, staking, and governance. The catch is that only one of those pillars is unquestionably “always on” at scale: fees. Governance participation tends to be episodic. Staking can be meaningful, but it is also an inflation channel that creates its own sustainability debt if real fee revenue does not eventually carry the security budget.

So the strategic read is simple. Starknet is trying to convert STRK from a political token into a resource token. The design now has to survive its own success. Every subsidy that helped bootstrap usage becomes harder to justify once fees are paid in the same asset that has large scheduled unlocks and an explicit minting path.

Supply: 10B initially, heavy allocations, and a long unlock runway

10,000,000,000 STRK were initially created by StarkWare in May 2022 and minted onchain on November 30, 2022. The docs also state that supply is not structurally fixed over time because new tokens can be minted through protocol mechanisms tied to staking and related rewards.

On March 5, 2026, CoinGecko reports circulating supply of 5,488,301,918 STRK, with total supply of 10,000,000,000 and a listed max supply of 10,000,000,000 on the STRK supply page. That “max supply” presentation is a market-data convention. It sits in tension with Starknet’s own documentation that describes future minting beyond the initial 10B.

The initial allocation plan is published in Starknet’s protocol docs as percentages of the initial 10B.

The lock-up mechanics that matter most for market structure apply to the combined Investors and Early Contributors buckets. The docs specify that up to 64M STRK (0.64% of total supply) unlocked on the 15th of each month from April 15, 2024 through March 15, 2025, totaling 768M STRK by March 15, 2025. They further specify up to 127M STRK (1.27% of total supply) unlocking on the 15th of each month from April 15, 2025 through March 15, 2027, totaling 3.048B STRK by March 15, 2027.

During this lock-up period, the docs state holders cannot transfer, sell, or pledge their locked STRK, while delegation of voting is permitted. The same page states that, as of its last update, staking with locked tokens is not permitted.

Utility and fiscal flows: fees route to the sequencer, not to holders

Starknet’s fee flow is unusually explicit at the protocol level. The fee flow rules describe how the fee for a transaction is charged atomically on L2 by injecting a transfer of STRK from the transaction submitter to the sequencer as receiver. The same docs state Starknet does not currently implement burning of fees, and all fees charged are received by the sequencer.

The STRK overview page adds a practical detail that matters for post-incentive equilibrium. It notes that a portion of fees paid in STRK may be converted to ETH by the receiving sequencer to cover Ethereum L1 gas costs, which must be paid in ETH. That implies a built-in “sell pressure” valve that can appear even under pure organic usage.

Starknet also chose to make STRK the sole native fee token while leaning on paymasters as the compatibility layer for users who do not want to hold STRK. In related Starknet proposals, transaction version 3 is framed as facilitating fee payment in STRK as the only native fee token, while older versions support fee payment in ETH, with paymasters positioned as the path to “multiple fee tokens” once ETH is no longer native.

That’s coherent engineering. Economically, it cuts both ways. Making STRK the only native fee token increases structural demand. Normalizing paymasters reduces the need for end users to ever hold STRK. If your long-term token thesis depends on “everyone needs STRK for gas,” paymasters are the leak in the bucket.

On the subsidy side, Starknet’s first big distribution program, as described in the Provisions program post, stated it would distribute more than 700M STRK to nearly 1.3M addresses, with claiming starting on February 20, 2024 and ending no later than June 20, 2024. That post also includes an update stating that the Provisions Program has been discontinued.

Discontinuations are not inherently bad. They do signal policy discretion. For sustainability modeling, discretion increases parameter uncertainty. You can’t cleanly project future net issuance, future rebates, or future “growth spend” when prior flagship programs are explicitly framed as changeable experiments.

Staking and inflation: the security budget is minted first, earned later

Starknet’s protocol docs state that staking is live and validators are required to attest blocks. The staking documentation describes a phased rollout and states the protocol is currently in its second of four phases on both Sepolia and Mainnet.

Mechanically, Starknet’s staking protocol sets a minimum validator stake of 20,000 STRK on mainnet. It also supports delegators and defines validator responsibilities around attestations, including epochs and an attestation window. Exiting carries a 7-day withdrawal security lockup on mainnet.

For a comparison with another emissions-driven model, see our Helium tokenomics review.

The economically decisive line is the rewards source. The staking docs state that staking rewards are issued in newly minted STRK tokens. They also publish an effective inflation coefficient of 4% for mainnet. Separately, StarkWare published a minting proposal that frames an inflation cap of 4% annually and provides a specific minting curve design philosophy for PoS.

Starknet also extends staking power beyond STRK. The staking docs state that starting Q3 2025, BTC holders can stake tokenized BTC representations (“wrappers”) on Starknet and earn rewards in STRK. The same page defines a BTC weight of 0.25 in staking power. This is a meaningful design choice because it increases the set of actors who can compete for minted STRK without necessarily being long-term aligned STRK holders.

From a long-term sustainability lens, staking is where Starknet’s token economy either matures or gets stuck. In the mature state, fees become a real budget and minting becomes marginal. In the stuck state, minting is the budget and fees are a rounding error. The protocol can function in both modes. Only one is durable without constant narrative support.

Governance: onchain voting exists, but delegation concentration is the real control surface

Starknet’s governance stack is real. The project states that its first-ever governance vote on mainnet occurred on September 10-13, 2024 and that governance is powered by Snapshot X using Herodotus Storage Proofs. Snapshot X is described by Starknet as publishing and verifying space settings, proposals, and votes onchain on Starknet while using storage proofs to verify balances across chains.

The mechanics also introduce friction that affects voter participation and, therefore, legitimacy. Starknet’s governance vote post states that participants must delegate their voting power before the vote, and that only people who held STRK and delegated before the snapshot date could vote.

Historically, Starknet described governance voting via a wrapped token. The Starknet FAQ states governance uses a unique token vSTRK, created by locking STRK at a 1:1 ratio, and that vSTRK can be converted back to STRK. It also states that vSTRK on Starknet and STRK on Ethereum can be delegated. More recently, the Governance Hub interface states that vSTRK is being deprecated, that it will no longer have voting power, and that Starknet is transitioning to a governance system using staked STRK tokens.

The other critical control surface is Foundation-influenced delegation. A Starknet Foundation governance forum post states it was rolling out a 3-tier delegate system to redistribute 1.7B STRK voting power, including tier allocations of 700M, 600M, and 400M STRK voting power across tiers. This is not automatically “bad governance.” It is an explicit admission that turnout and representation are not self-sustaining yet, so voting power is actively curated to keep governance functioning.

For tokenomics, that implies governance risk is not theoretical. It is operational. Fee policy, minting policy, staking parameters, and even which token is required for gas have already changed over time through coordinated upgrades. If you are valuing STRK as a long-duration asset, governance credibility and constraint design matter at least as much as near-term TVL.

Risk analysis: the post-incentive equilibrium is still being negotiated

Starknet’s token economy is increasingly “real” in the sense that it has binding protocol flows: users pay fees in STRK, fees go to the sequencer, and staking can mint rewards. The unresolved part is whether these flows converge to a stable equilibrium once the ecosystem stops leaning on grants, rebates, and periodic distributions.

If you want a framework for evaluating those trade-offs, our tokenomics methodology page explains the standards we use when stress-testing design assumptions.

Top 3 risks

  1. Dominant risk: Monetary and fiscal credibility breaks under competing stakeholders.

    Trigger: Governance approves higher effective inflation, expands staking reward scope, or increases recurring “growth spend” from reserves in response to market stress, validator lobbying, or ecosystem slowdown.

    Mechanism: Starknet’s security budget is explicitly mint-powered today because staking rewards are issued as newly minted STRK. Meanwhile, the protocol does not burn fees and routes fees to the sequencer. In that configuration, “real yield” only emerges if fee revenue becomes large enough and credibly shared, or if inflation remains tightly bounded. If either assumption weakens, STRK becomes a political object again. Stakeholders rationally push for policy that protects their cash flows, even if it harms long-run token purchasing power. This is the standard failure mode of subsidy-first systems. They struggle to cut spending when the narrative shifts from growth to maintenance.

    Starknet also has a built-in tension created by fee-token design. Fees became STRK-only in v0.14.0, which increases structural demand, but Starknet proposals emphasize paymasters as the way to preserve multi-asset UX. If paymasters become dominant, end-user demand to hold STRK can drop even while the chain grows. That pushes the system toward “validators get minted STRK, paymasters buy STRK to pay fees, sequencer sells STRK for ETH,” which is functional but not obviously value-accretive for passive holders.

    Who bears it: Long-only STRK holders, delegators who receive inflation-denominated rewards that may not compensate for dilution, and application teams whose runway depends on predictable subsidy regimes.

    Measurable indicators: Rising minted STRK relative to fee revenue, persistent sell pressure around unlock dates, repeated governance proposals that expand rewards or rebates, and increasing delegation concentration driven by Foundation-curated voting power.

    For ongoing monitoring, we publish periodic crypto research that can help ground these indicators in data instead of narratives.

    Why this is dominant: It is the risk that can silently absorb every other risk. If the monetary and fiscal regime is not credible, “good tech” does not rescue the token economy. The chain can still win users and developers. STRK can still underperform as a store of value because the system’s first instinct under stress will be to mint and spend, not to retrench. Starknet’s own public history includes large-scale community distribution plans and explicit program discontinuations. That pattern is consistent with experimentation. It is also consistent with policy instability.

  2. Unlock overhang collides with “STRK-only fees.”

    Trigger: Monthly investor and early contributor unlocks continue through March 15, 2027 while network usage is not growing fast enough to absorb incremental supply.

    Mechanism: STRK-only fees force applications and users to source STRK somewhere. If a meaningful portion of unlocked supply is sold into the market, that supply becomes the marginal source of “gas liquidity.” The network works, but token price becomes the shock absorber for both usage volatility and unlock schedules. Paymasters reduce user pain, but they also concentrate fee purchasing into specialized entities that can optimize execution and hedging.

    Who bears it: Retail holders and ecosystem participants paid in STRK grants who have expenses in USD or ETH, plus delegators whose rewards are paid in STRK.

    Measurable indicators: Price weakness around the 15th of each month, widening spread between circulating supply and staked supply, and persistent reliance on Foundation distributions to maintain activity.

  3. Governance capture via delegated voting power becomes normalized.

    Trigger: Low participation persists, leading to outcomes dominated by a small delegate set and by Foundation-redistributed voting power.

    Mechanism: Voting requires delegation before the snapshot, which is friction that naturally reduces turnout. When governance is repeatedly “saved” by curated delegation, the system can drift into a stable but narrow decision-making equilibrium. Policy becomes predictable for insiders and unpredictable for everyone else. That is a direct hit to long-term parameter stability for minting, staking, and fee policy.

    Who bears it: Passive holders, smaller builders without governance bandwidth, and users if governance decisions prioritize validator or operator economics over UX.

    Measurable indicators: Delegate concentration metrics, repeat wins by the same voting blocs, and governance changes that increase complexity or reduce contestability, such as shifting voting power from vSTRK to staked STRK without clear, broadly validated constraints.

If you are doing serious scenario work on STRK and need help stress-testing assumptions around unlock absorption, staking inflation, and governance-driven parameter risk, this is where disciplined modeling beats vibes. A short engagement with a team that does tokenomics consulting can be justified purely by avoided blind spots, not by “optimizing APR.”



This article is part of our Tokenomics Deep Dive series.