GT Inflation Analysis · September 2026 · Mixed flows, supply roughly steady

GateToken (GT) is the exchange token of Gate and the staking and gas asset of GateChain, and over the 90 days to Sep 5 2026 the MrNasdog Pressure Framework reads 0.09M GT of sell pressure against zero buy pressure on a circulating base of 106,616,440 GT, for a net of +0.08%. Gate did burn 2,570,063 GT on Jul 4 2026, lifting the lifetime burn to 189,947,220 GT — about 63% of the original 300,000,000 — and that burn is completely real. Every burned GateToken came out of a frozen reserve wallet that received nothing at all across the window, and that reserve is precisely the slice of supply the circulating count already excludes, so the ceiling fell and the tradable GT float did not move by a single token. Meanwhile GateChain paid its consensus nodes in newly issued GT. GateToken is slightly inflationary, not deflationary.

The verdict, in one paragraph

For the 90-day window ending Sep 5 2026, the Pressure Framework reads GT at +0.08% net: sell pressure of 0.09M GT, buy pressure of 0, on a circulating base of 106,616,440 GT. The next 90 days read +0.08% as well, because the only mechanism that adds GateToken — GateChain consensus issuance — is continuous, and the only mechanism that removes it, the quarterly burn expected around Oct 2026, draws on the excluded reserve. Our supply monitor reads -0.035% for the same window, a gap of 0.12 percentage points, comfortably inside the half-point tolerance, so no monitor-gap flag is raised on this build and the two readings agree that GT's tradable supply is close to flat. The right label for GateToken is an exchange token whose burn shrinks the ceiling while its own chain slowly grows the float.

Sell pressure: where new GT comes from

Sell #1, protocol inflation, is 0.09M GT, and it is the row that decides the sign of this page. GateToken is almost always described as fixed at 300,000,000, and on Ethereum that description is exactly right: the GateToken contract is a 2019 ERC-20 whose entire function list was read out of the deployed bytecode one selector at a time, and there is no mint of any kind in it. The supply read was 300,000,000.000000 at both ends of this window and it always will be. GateChain is the part that gets missed. It is a proof-of-stake chain, and it pays its consensus nodes in newly issued GT drawn from a 30,000,000 GT allocation of which Gate's own supply page shows 8,748,062 GT mined and 21,251,938 GT still unmined — two figures that add to exactly thirty million. That issuance was sized two independent ways: against Gate's own earlier published issuance figure the counter has advanced about 76,500 GT per 90 days, and taken per block it works out at 0.397 GT a block across 257,532 GateChain blocks in a quarter, or about 102,200 GT. Two angles, twenty-nine percent apart, so the page ships the middle.

That second angle needed a measured block interval rather than a published one. GateChain's realised interval over a hundred-thousand-block sample this week is 30.19 seconds, against a lifetime average nearer 9.9 seconds since the 2019 mainnet — the chain has slowed roughly threefold as activity moved to Gate Layer. Because the consensus reward is paid per block and nothing in the protocol rescales for a slower block, the lifetime average issuance rate of about 312,600 GTa quarter is a rate that no longer exists, and using it would have overstated GateToken's inflation by more than three times.

Sell #2, vesting unlocks, is 0. GateToken finished its distribution in 2019 and has never had a vesting contract, a cliff calendar or an escrow to read. Gate's own supply split is the cleanest proof: circulating 97,801,549 plus burned 189,947,219 plus frozen 12,251,232 comes to exactly 300,000,000, with no vesting bucket anywhere in it. The one unlock-tracker page that claims otherwise for GT prints a next release of 283.7 billion tokens for a token whose entire genesis was 300 million, which is not a schedule but a fabrication, so it is discarded and the question settled against the wallets themselves.

Sell #3, foundation and unscheduled unlocks, is 0 — the ability to sell is not the same as a decision to sell — but the overhang is listed rather than waved away, and for GateToken it is unusually large. The frozen 2019 reserve holds 12,251,232 GT and only ever pays the burn. The executor wallet that routes each burn holds 12,620,975 GT and was identical at both ends of the window. Three further Gate wallets hold 52,399,955 GT, 24,323,323 GT and 7,038,973 GT; two of them did not move a GateToken all window and the third grew by 231,279 GT, an inflow rather than a sale. Sell #4 is 0: GT has no bankruptcy estate, no trustee and no court-ordered distribution, because Gate is a going concern.

Buy pressure: where new GT goes

Buy #1, programmatic buyback, is 0, and this single row is the whole GateToken story. The quarterly burn is real, it fired inside this window, and it settles on Ethereum: 2,570,063 GT reached the burn address on Jul 4 2026 in one transaction, announced two days later, taking the burn address from 187,377,156 GT to 189,947,220 GT. Both supply surfaces were read at both ends of the window, because either one alone gives the wrong answer here — total supply never moved at all, and a build watching only total supply would have reported no burn on the largest burn programme in exchange-token history. What decides the row is who sent the coins, because a treasury-funded burn and a market-funded burn look identical on a balance chart and are opposite in the ledger. Traced by sender, the frozen 2019 reserve took in nothing across ninety days and paid out exactly 2,570,063 GT, reimbursing the executor wallet the same day the burn cleared.

The subtraction that settles it is one line. Published total supply minus published circulating supply is 12,251,290 GT; the reserve's balance is 12,251,232 GT. Those are the same number to within fifty-eight GateTokens. The entire non-circulating bucket is that one wallet, and it is the wallet the burn is paid from, so the burn lowered GateToken's ceiling and removed no tradable supply. Booking it as absorbed float would have printed -2.41% of deflation that no GT holder experienced, and the measured float agrees: circulating supply moved by about -37,000 GT across the whole quarter.

Buy #2, protocol fee burn, is 0, and it was measured rather than assumed. Sweeping every transfer into the burn address across 645,607 Ethereum blocks returned exactly one — the quarterly burn — so there is no continuous destruction on the ledger of record, and the swept total closed against an independently read balance to the last decimal. GateChain does burn its own base fee, but the chain runs at about a tenth of a transaction per block and the sampled burn is roughly 5 GT across the quarter. Gate Layer routes its base fee to a fee vault rather than destroying it. Buy #3, foundation buy, is 0: no open-market GateToken purchase by Gate has been disclosed, and the reserve that funds the burn received nothing at all. Buy #4, new long-term lock, is 0 — no new lockup was created, and the only long-term reserve GateToken has got smaller this quarter, not bigger.

Foundation and overhang

GateToken's overhang is concentrated in six wallets that between them hold almost every GT that exists on Ethereum. The frozen 2019 reserve at 12,251,232 GT is the only one outside the circulating count, and at the current burn rate of about 2.6M GT a quarter it holds roughly four to five more quarters of burn fuel before Gate must fund burns from somewhere else — the single most important thing to watch on this token. The burn executor at 12,620,975 GT sits inside the circulating count and behaves as a pass-through, unchanged at both ends of the window. The consolidated Gate cold wallet at 52,399,955 GT is the largest genuine overhang, roughly half the tradable float, and it did not move a GateToken across the quarter; a second company wallet holds 24,323,323 GT and grew by 231,279 GT; a third holds 7,038,973 GT and was flat. Beyond those, the unmined 21,251,938 GT of the GateChain consensus allocation is a slow-release overhang that reaches the market a block at a time rather than in a cliff. If any of these balances falls between refreshes, the outflow enters Sell #3 at the next refresh.

How GT compares to other exchange tokens

The natural comparison class is exchange tokens with programmatic burns, and GateToken sits at an unusual corner of it. The dominant model in that class buys tokens on the open market with a share of exchange revenue and then destroys them, which removes tradable float and shows up as genuine buy pressure in a framework like this one. GateToken's burn is sized by that same revenue rule, but the tokens it destroys are drawn from a frozen internal reserve rather than bought back from an order book, so the mechanism reads as a ceiling reduction rather than a float reduction. Both are real deflation on total supply; only one of them competes with sellers for the coins people can actually trade. The difference is invisible on a supply chart and decisive in the ledger, which is why the sender trace matters more than the burn headline.

The second structural difference is that GateToken is not only an exchange token. Most exchange tokens are pure ERC-20s with no issuance of their own; GateToken is also the native staking asset of GateChain, a proof-of-stake Layer 1, and of Gate Layer above it. That gives GT a small permanent issuance leg that a pure exchange token does not have, funded from a 30,000,000 GT allocation with 21,251,938 GTstill to come. Against a hard-capped chain with a halving schedule, GateToken has no protocol-encoded scarcity clock at all: its supply path is set by two corporate decisions — how fast the reserve is burned and how long the consensus allocation lasts — rather than by code that cannot be changed. And against an uncapped continuous-emission Layer 1, GateToken's issuance is tiny, well under a tenth of a percent a quarter, which is why the page reads as roughly steady rather than clearly inflationary.

What to watch in the next 90 days

First, the Q3 2026 quarterly burn, expected around Oct 2026 on the pattern set by Apr 2026 and Jul 4 2026: the number to check is not the burn size but the sender, because a burn funded from anywhere other than the frozen reserve would flip Buy #1 from zero to real buy pressure overnight. Second, the reserve balance itself at 12,251,232 GT — four or five more burns and it is empty. Third, the burn executor's 12,620,975 GT, which already sits inside the circulating count and has no announced destination beyond the burn address. Fourth, GateChain's block interval, currently 30.19 seconds and slowing, since a block-indexed consensus reward means a slower chain issues less GateToken. Fifth, any Gate announcement moving GT issuance or fee burning onto Gate Layer, which would change which chain the framework has to measure.

Summary

GateToken (GT) reads +0.08% net supply growth over the 90 days to Sep 5 2026 and the same again forward, on a circulating base of 106,616,440 GT — slightly inflationary, not deflationary, despite the most famous burn programme in the exchange-token category. The structural mechanism is a split one: the Ethereum contract can never mint another GateToken, while GateChain issues about 0.09M GT a quarter to its consensus nodes, and the 2,570,063 GT burned on Jul 4 2026 came from a frozen reserve that the circulating count already excludes, so it shrank the ceiling and not the float. The key risk is that the same reserve is finite at 12,251,232 GT, about four or five more quarterly burns, after which Gate must either fund burns from the market — which would turn a zero buy row into a real one — or stop. The ceiling is 300,000,000 GT on Ethereum plus a 30,000,000 GT consensus allocation that is 21,251,938 GT from being spent.

MrNasdog Pressure Framework analysis of GT, Metric 1 — Inflation. Data + explanation only. Not financial advice. Updated Sep 5 2026.