JUP Inflation Analysis · September 2026 · Mixed flows, supply roughly steady
Jupiter cannot mint a single new JUP — the mint authority on the JUP token is renounced on-chain, and the February 2026 net-zero governance vote ended the vesting schedule outright. What still moves JUP supply is a pair of opposing flows, and over the last 90 days the larger one pointed the wrong way for holders: Jupiter's community wallet released 48.09M JUP of quarterly staking rewards, while the Litterbox Trust bought and parked 32.79M JUP with half of Jupiter's protocol revenue. The MrNasdog Pressure Framework reads JUP at +0.46% net over the last 90 days on a circulating base of 3,320.31M JUP, and projects the same +0.46% for the next 90 days, because one more quarterly rewards payout falls inside that window.
The verdict, in one paragraph
For the 90-day window ending Sep 5 2026, the framework reads JUP at +0.46% net: 48.09M JUP of sell pressure against 32.79M JUP of buy pressure, on a circulating base of 3,320.31M JUP. Our supply monitor reads the same window at +0.01%, a gap of 0.45 percentage points — inside the framework's half-point tolerance, so no monitor-gap chip ships on the JUP overview. That the two nearly agree is itself informative: both of the big flows in this window run through wallets a supply count already treats as circulating, so they largely cancel in that series while the framework sees them separately. The honest label for Jupiter today is a revenue buyback that its own rewards programme still outruns. Two things about JUP genuinely are locked shut — issuance and vesting — and the residual growth comes from neither.
Sell pressure: where new JUP comes from
Nowhere, in the strict sense. Protocol inflation is 0: a read of the JUP mint on Solana returns a null mint authority and a null freeze authority, so no party — not the team, not the DAO, not a future governance vote — can create another JUP. Total supply stands at 6,862.43M JUP and can only fall. Vesting unlocks are 0 for the same structural reason a calendar can be closed: the net-zero governance vote of Feb 22 2026 did not pause the schedule, it ended it. The last team unlock was Jan 27 2026, the founding-stakeholder pool was settled with a single accelerated tranche of 77.97M JUP on Feb 25 2026 in place of its remaining monthly releases, and the postponed airdrop allocations — a 1B reserve plus 200M each for 2026 and 2027 — were cancelled into custody with no release date at all. Unlock trackers now describe JUP as fully unlocked with nothing upcoming. A retired monthly tranche of about 53.47M JUP is still republished by some listings against fresh dates; it is dead, and this page does not book it.
The 48.09M JUP that the framework does book sits under Foundation and unscheduled unlocks, and it is real. Jupiter's community wallet paid out its quarterly staking rewards: 50.00M JUP left that wallet on Jul 7 2026 and Jul 16 2026 into two claim contracts, 1.91M JUP flowed back from the previous round, and both claim contracts have since emptied to stakers with only 0.49M JUP ever clawed back. The equivalent payout ran on Apr 8 2026 and again on Jul 8 2026, which is why the next 90 days carries one more. A second large movement in the window is deliberately not booked: 90.00M JUP left one of Jupiter's team vaults, but every hop stayed inside Jupiter's own multisig infrastructure and all nine receiving wallets still hold the full amount today — a change of custody, not coins reaching the market. Long-term locked or bankruptcy is 0: Jupiter has no bankruptcy estate and no trustee distribution.
Buy pressure: where new JUP goes
The whole buy side is one mechanism. Jupiter directs half of its protocol revenue — revenue, not the larger gross-fee line — into open-market JUP purchases held by the Litterbox Trust under a three-year lock. Read straight off the chain, the trust went from 132.73M JUP on Jun 7 2026 to 165.52M JUP on Sep 5 2026: 32.79M JUP bought in 90 days, rising at every checkpoint in between and never once falling. The arithmetic on the split closes cleanly. Jupiter took $15.07M of protocol revenue over that window; half of it is $7.54M, which against the measured purchase implies an average price of $0.23 per JUP — right on the range JUP traded through the summer. Run the same 50% off gross fees instead and it would imply nearly 120M JUP, more than three times the truth; that distinction is the single most common way this mechanism gets misread.
Protocol fee burn is 0, and this matters more than it sounds, because the Litterbox Trust parks JUP rather than destroying it. Both surfaces that could reveal a burn were read at each end of the window: the unspendable address holds effectively nothing, and total supply did not move. That supply figure is not a dead field — an SPL burn writes it directly, and it has moved before, from a 10B genesis cap down to 7B after the community supply reduction and then to today's 6,862.43M JUP after the roughly 134M JUP burn holders voted through in Nov 2025. It simply did not move here. So the bought-back JUP still exists, and it reappears on the sell side of the ledger as a tracked overhang. Foundation buy is 0 — no discretionary purchasing is disclosed beyond the revenue buyback — and new long-term lock is 0, because the trust's three-year lock is already inside the buyback row and is not counted twice.
Foundation and overhang
Jupiter's overhang is larger than its float. About 3,542M JUP sits outside the circulating count in three reserve wallets — 1,700.0M, 1,682.7M and 159.9M — and this build read all three at both ends of the window with no transaction in between: they did not move at all. Those wallets hold the cancelled airdrop allocations and the ended team and founding-stakeholder blocks, none of which now carries a release date. Inside the float there are four more identified Jupiter wallets worth watching, all read on-chain this window: the Litterbox Trust at 165.5M JUP, the community wallet at 223.4M JUP that paid this quarter's rewards, a team vault at 191.6M JUP, and a team hot wallet at 126.4M JUP that did not move once in 90 days. The nine wallets that received the internal 90.00M JUP transfer hold 90.10M JUP between them today.
Every one of these is refreshed by a direct on-chain read at each rebuild, and the rule is the same for all of them: if any of these balances falls between refreshes, that outflow enters the Foundation and unscheduled unlocks row at the next refresh. That is precisely what happened this build — the community wallet fell, so it is booked. It is also why the Litterbox Trust appears on both sides of the ledger at once. A three-year lock is not a burn, the JUP inside it is still JUP, and a supply count still sees those coins in the circulating float, so the framework credits the purchase on the buy side and carries the balance as an overhang on the sell side.
How JUP compares to other DeFi application tokens
Judged purely on issuance, JUP sits at the top of its class. Most DeFi application tokens still run a live emission — a liquidity-mining programme, a staking subsidy, or a vesting calendar with years left on it — and their sell side is dominated by supply that did not exist a quarter ago. Jupiter has none of that. The mint authority is renounced, which is stronger than a hard cap because a cap can be lifted by governance while a renounced authority cannot be restored, and the vesting calendar was not paused but cancelled by vote. Against a token like a typical exchange or aggregator token that offsets ongoing emission with a periodic burn, Jupiter is the rarer shape: nothing is being created, and the residual supply movement is entirely redistribution of coins minted years ago.
Where Jupiter separates itself further is the size of the fee engine relative to the token. On the last 30 days of activity, Jupiter's gross trading fees annualise to roughly 31% of its market capitalisation, and the narrower protocol-revenue line — the one the buyback actually spends — annualises to about 10.5%. Taking the conservative basis, protocol revenue against market cap, that ratio is at the very top of the DeFi field; a majority of large application tokens sit in low single digits or below one percent, and many carry no protocol revenue at all. It is worth being precise about scale rather than reaching for a superlative: Jupiter took $18.4M of gross fees in the 30 days to Sep 5 2026, of which $6.2M was protocol revenue.
The comparison that decides the reading, though, is buyback-and-burn versus buyback-and-hold. Exchange tokens that burn what they buy retire the supply permanently and the float ratchets down. Jupiter buys and locks, and its own history shows the difference is not academic: the 134M JUP destroyed in Nov 2025 came out of this same trust and required a separate governance vote to make it a burn. Until such a vote happens again, every JUP the trust buys is parked, not gone — and this window the parking was smaller than the rewards released.
What to watch in the next 90 days
Four things move this reading. First, the next quarterly staking-rewards payout, expected around Oct 8 2026 on the rhythm set by Apr 8 2026 and Jul 8 2026; it is the single largest input on the sell side and the reason the next-90-day projection is not negative. Second, the buyback pace, which tracks Jupiter's protocol revenue directly — a sustained revenue recovery would lift the buy side above the rewards release and flip the net without any governance action at all. Third, a live forum proposal posted May 30 2026 to raise the buyback share from 50% to 70% of protocol revenue; it has not been put to a vote, and the vote portal's last completed proposal of any kind remains the Feb 2026 net-zero vote. Fourth, any proposal to burn rather than hold the trust's balance, which would convert 165.5M JUP of tracked overhang into permanently retired supply.
Summary
Jupiter is a token that cannot be printed sitting inside a protocol that keeps redistributing it. The mint authority on JUP is renounced and the vesting schedule was cancelled by governance in Feb 2026, so nothing new enters the market — but Jupiter's community wallet released 48.09M JUP of quarterly staking rewards in the last 90 days against a 32.79M JUP revenue buyback, leaving the framework at +0.46% net and projecting the same for the next 90 days on one further rewards payout. The key risk is not dilution, which the renounced mint authority rules out, but distribution: 3,542M JUP of reserves sit outside the float with no schedule, and the 165.5M JUP the buyback has accumulated is locked for three years rather than destroyed. The ceiling is real and getting no higher — total supply is 6,862.43M JUP and can only fall — though the wider market has not caught up: a major price listing still publishes the retired 10,000Mgenesis cap as JUP's maximum supply.
MrNasdog Pressure Framework analysis of JUP, Metric 1 — Inflation. Data + explanation only. Not financial advice. Updated Sep 5 2026.