GRT is an inflationary work token: the protocol issues roughly 3% new supply every year to pay Indexers, and it burns tokens through delegation, curation, and query fees to push the other way. There is no fixed maximum supply, so whether GRT expands or contracts in a given period comes down to one balance, how much burning activity offsets the fixed issuance. Understanding that balance is the key to understanding what backs the token.
Most token guides stop at "GRT is used for staking and queries." That misses the mechanism that actually moves supply. This breakdown walks through issuance, rewards, and burns as one connected system, because that is how they behave.
How GRT Supply Works
GRT launched with an initial supply of 10 billion tokens after its October 2020 sale at $0.03 each. As of July 2026, total supply sits near 10.8 billion, with circulating supply close behind. The token has drifted upward over time by design, since the protocol mints new GRT continuously to reward the participants who keep data available.
The number that matters is the issuance rate. The protocol targets around 3% new GRT per year, distributed as indexing rewards. That is a deliberate subsidy, paying Indexers to index and serve subgraphs even before enough query fees exist to cover their costs. Without it, a new subgraph would sit unindexed until demand appeared, which is the classic bootstrapping problem for any data network.
Where Indexing Rewards Come From
The 3% issuance flows to Indexers based on how much they stake and which subgraphs they index. To claim rewards, an Indexer submits a Proof of Indexing, a cryptographic attestation that it actually processed the data correctly. Serve bad data and that stake can be slashed, which is what makes the reward honest rather than a passive payout.
Delegators plug into this same reward stream. By delegating GRT to an Indexer, they lend that Indexer more stake to work with and take a share of the resulting rewards, minus the Indexer's cut. Curators sit slightly apart: they signal GRT on subgraphs they judge valuable and earn a portion of the query fees those subgraphs later generate. Each role earns from a different slice of the same economic pie, and each one requires putting GRT to work.
The Burn Mechanisms
Issuance is only half the story. The Graph burns GRT at three points, permanently removing those tokens from supply. The table below shows each one.
| Action | Burn | What it removes |
|---|---|---|
| Delegating GRT to an Indexer | 0.5% delegation tax | Burned on every delegation |
| Signaling on a subgraph (curation) | 1% curation tax | Burned when Curators signal |
| Query fees paid by consumers | 1% of all query fees | Burned per query settled |
None of these burns is large on its own. Their weight depends entirely on volume. Heavy delegation, active curation, and high query demand all shrink supply, while quiet periods let the 3% issuance dominate. The protocol also enforces a 28-epoch unbonding period, roughly 26 to 28 days, before Delegators can withdraw, which slows reflexive exits and keeps stake committed.
How Supply Dynamics Shape the Token
Put issuance and burns side by side and the investment question becomes concrete. GRT holders are effectively long a bet that network activity will grow enough for burns to catch and eventually outpace the 3% issuance. That is the same variable driving most GRT price prediction scenarios, since a token that keeps diluting struggles to appreciate no matter how strong the underlying usage looks.
The current picture is mixed. The network has served over a trillion cumulative queries and query fees on Arbitrum hit records in 2025, yet net supply has kept expanding because burn volume has not consistently exceeded issuance. Usage growth is real, but the tokenomics reward patient holders only if that growth translates into paid, high-value queries rather than cheap or subsidized ones.
What Horizon Changes for Tokenomics
The Horizon upgrade, live since December 2025, reworks how rewards get earned. A Rewards Eligibility Oracle ties indexing rewards to proof of actual value delivered, aiming to stop rewards from flowing to Indexers who stake but serve little real demand. Direct Indexer Payments let consumers and chains compensate Indexers outside the standard query-fee path, adding flexible demand-side incentives.
The intent is to make issuance productive. If rewards track genuine service, the 3% subsidy funds real capacity instead of passive farming, and the burn side grows as paid demand does. Whether that rebalances GRT toward net-deflationary territory is the open question the 2026 roadmap is trying to answer, and it is worth tracking through official reporting from The Graph's documentation and independent coverage on CoinMarketCap.
Frequently Asked Questions
Does GRT have a maximum supply?
No. GRT has no hard cap. The protocol issues roughly 3% new tokens annually as indexing rewards and burns tokens through delegation taxes, curation taxes, and a share of query fees. Net supply change in any period depends on whether burn volume exceeds issuance.
How much GRT is burned?
Three burns apply: a 0.5% tax when Delegators delegate, a 1% tax when Curators signal on a subgraph, and 1% of every query fee. The total burned depends on network activity, so higher delegation, curation, and query volume remove more GRT from circulation.
Is GRT inflationary or deflationary?
GRT is inflationary by default because of the fixed 3% annual issuance. It can trend deflationary in periods when burns from queries, curation, and delegation exceed that issuance, but sustained deflation requires high, consistent network demand that has not yet been the norm.
What GRT's Token Design Means for Holders
GRT's tokenomics reward network growth and penalize stagnation with dilution. The 3% issuance is a bet the protocol makes on itself, subsidizing data availability today in exchange for the demand it hopes to attract tomorrow. For holders, the entire thesis reduces to whether paid query volume and curation can outrun that subsidy over time.
Horizon's reward reforms are the most serious attempt yet to tilt that balance toward productive issuance and heavier burns. They do not remove the risk, they just make the token's fate depend more directly on real usage, which is where an infrastructure token should sit.
Trade GRT on the spot market or open a leveraged position with GRT perpetual futures on LeveX. Browse the Crypto in a Minute series for more token guides.