LINK Inflation Analysis · August 2026 · Supply growing, projected to keep growing
Chainlink has a hard 1 billion LINK cap and mints nothing, yet the circulating supply of LINK still grows, because Chainlink releases LINK out of a non-circulating reserve at a published rate. Reading all 33 of those reserve wallets directly at both ends of the last 90 days shows them falling from 272.9M to 251.9M — exactly 21M LINK reached the float, in a single release on Jun 19 2026 — against 1.8M LINK absorbed and held by the Chainlink Reserve. The MrNasdog Pressure Framework reads +2.57% net over that window; the supply monitor reads +2.91%, a gap of only 0.34 percentage points, inside tolerance and with no data-conflict flag.
The verdict, in one paragraph
For the 90-day window ending Aug 24 2026, the MrNasdog Pressure Framework reads LINK at +2.57% net on the realised view and +2.07% net on the forward view, driven entirely by reserve release rather than by any protocol mint. The supply monitor reads the realised change at +2.91% for the same window, so the gap is 0.34 percentage points — inside tolerance, and this build therefore ships no monitor-gap flag. Most of the residual is the Chainlink Reserve's 1.8M LINK absorption, which the upstream classifier still counts as circulating while the framework treats it as taken off the market. Both the realised and the forward reading sit inside the 1% to 3% growth band, so the inflation score stays at 1 on the 0–5 scale. The label for Chainlink: hard-capped, zero-mint, but structurally inflationary on the active float for as long as the reserve unwinds.
Sell pressure: where new LINK comes from
Sell #1, protocol inflation, is zero, and that is the defining fact about Chainlink tokenomics. LINK is a fixed-supply token on Ethereum: all 1 billion LINK were minted at launch, there is no mint function, no block reward and no staking emission. An on-chain total-supply read at both ends of this window returned the same 1,000,000,000 LINK, so not a single new token was created. Sell #2, vesting unlocks, is also zero: Chainlink has no vesting contract and no cliff calendar, and an independent unlock tracker lists LINK as fully unlocked with nothing upcoming. Sell #4, long-term locked or bankruptcy, is zero as well, because no bankruptcy estate, trustee schedule or court-ordered distribution applies to LINK.
That leaves Sell #3, Foundation and unscheduled unlocks, as the entire sell story, at 21M LINK over the last 90 days. Because Chainlink publishes the full list of its non-circulating wallets, this row is measured rather than estimated: the aggregate balance of those 33 addresses fell from 272,900,030 LINK to 251,900,030 LINK, a realised outflow of exactly 21,000,000 LINK. It moved in one cluster on Jun 19 2026, out of four reserve wallets, each of which sent a single test token before the bulk: 18,375,000 LINK went to a deposit address that held nothing at either end of the window and forwarded every unit onward to an exchange hot wallet, and 2,625,000 LINK went to a staking-rewards multisig. All four wallets were then unchanged from Jun 21 2026 to today, so there was no second release in July or August.
The forward projection for Chainlink is not an average of past flow and not the midpoint of recent firings — it is the published rule. Chainlink states that its token release schedule is currently 7% of the total supply per year, which on a 1 billion cap is 70,000,000 LINK a year. An independent twelve-month on-chain read of the same 33 addresses confirms it exactly: 321,900,030 LINK on Aug 24 2025 against 251,900,030 LINK today, a drawdown of 70,000,000 LINK to the token. The individual firings are irregular — the gaps between the last four were 70, 105 and 76 days, and no release calendar is published — so the rate is scaled to the framework's window rather than assumed to be one calendar quarter: 17.26M LINK for the next 90 days.
Buy pressure: where new LINK goes
Buy #1, the programmatic buyback, is the only active offset in the Chainlink ledger, at 1.8M LINK over 90 days, and it takes the form of the Chainlink Reserve. Through Payment Abstraction, on-chain service fees, off-chain enterprise revenue and half of the value recaptured by Chainlink SVR are converted into LINK and deposited into a reserve contract that accumulates behind a multi-day withdrawal timelock. Reading that contract directly gives 3,779,076 LINK on May 26 2026 and 5,581,719 LINK today, so 1,802,642 LINK was genuinely taken off the open market — a realised figure, not a headline rate. Chainlink's own quarterly review, an independent check on the chain read, reports the Reserve accumulating 1.44M LINK in the second quarter of 2026 to a total of 4.5M, which is what the on-chain balances bracket at the end of June.
The destination question matters more here than the amount, because the Chainlink Reserve holds. It does not burn and it does not stake: a full scan of the contract across this window found no outgoing transfer of any size, and Chainlink states it expects no withdrawals for multiple years. The pace is flat rather than accelerating — six consecutive fortnightly reads run at 18,740, 19,499, 29,446, 18,333, 18,476 and 15,430 LINK a day, with the most recent the lowest of them — so the next 90 days project to the same 1.8M LINK.
The remaining buy rows are zero. Buy #2, protocol fee burn, is zero because LINK has no burn mechanism at all: fees for oracle, data-feed and cross-chain services are routed to node operators and to the Chainlink Reserve, never destroyed. Buy #3, Foundation buy, is zero, with no discretionary open-market buying disclosed beyond the Reserve accumulation already counted; a listed company announcing its own LINK treasury does not belong in this row either, because those tokens are bought on the open market and already sit inside the circulating base. Buy #4, new long-term lock, is zero as well: the Chainlink community staking pool read 40,875,544 LINK at both ends of the window, identical to the decimal and at its published cap, and the node-operator pool moved by under 3,000 LINK. Chainlink staking v0.2 is still the live version with a 45M LINK total cap; a cap increase or a new staking version would be the event that puts a number in this row.
Foundation and overhang
The defining structural feature of LINK is its overhang. Roughly 251.9M LINK — about a quarter of the entire 1 billion cap — still sits outside circulation across the 33 non-circulating wallets Chainlink publishes, with no cliff calendar and no committed release dates, only the annual 7% rate. That pool is what every release draws from, it is by far the largest supply variable for the token, and it is re-read on-chain wallet by wallet on every rebuild. A second, smaller overhang is the staking-rewards multisig that receives a slice of each release: it held 8,430,000 LINK before the June firing, took in 2,625,000 LINK, and holds 10,055,000 LINK today as it pays stakers down at roughly half a million a month. On the other side of the ledger sits the Chainlink Reserve, the buyback destination, an on-chain accumulation contract holding 5,581,719 LINK today behind a multi-day timelock. Exchange deposit addresses are excluded from this enumeration because they belong to depositors, not to Chainlink, and the one that received the June release held nothing at either end of the window. If the balance of the non-circulating reserve, the staking multisig or the Chainlink Reserve falls between refreshes, that outflow enters Sell #3 at the next refresh.
How LINK compares to other capped-supply utility tokens
Chainlink belongs to the family of hard-capped, pre-minted utility tokens rather than to the continuous-emission layer-1 family. Against an uncapped proof-of-stake chain, which pays validators out of fresh issuance every block and therefore prints supply forever, LINK looks structurally superior: the cap is absolute, the mint function does not exist, and the ceiling on lifetime dilution was fixed at launch. Against a halving-model chain with a hard cap, the comparison inverts. A halving chain's remaining issuance is mechanical and decelerating on a schedule enforced by code; Chainlink's remaining 251.9M LINK is released at the discretion of the issuer, and the 7% per year rate is a policy statement rather than a protocol constraint — it can be changed without an upgrade, and the individual release dates are known only after the transfer lands.
The closest structural analogue is an exchange token or an enterprise token with a large treasury float and a revenue-funded buyback. Both accumulate value by routing fees back into the token, and both carry a treasury overhang that dwarfs the buyback. The critical mechanical difference is destination: an exchange token typically burns what it buys, permanently shrinking supply, while the Chainlink Reserve holds what it accumulates. Held tokens can, in principle, come back; burned tokens cannot. That single design choice is why the Chainlink Reserve, at under 2M LINK a quarter, offsets rather than reverses a 21M LINK release, and why the framework reads Chainlink as inflationary on the float despite a genuinely fixed cap.
The offsetting argument, which the mechanism supports, is direction of travel. Reserve release is a finite process — the non-circulating pool only shrinks, and at 7% of the cap a year it has roughly three and a half years of full-rate release left in it — while Reserve accumulation is tied to network revenue and has climbed from a standing start in Aug 2025 to 5.58M LINK a year later. If revenue keeps compounding while the release rate holds at 70M LINK a year, the two lines converge. That crossover has not happened, and on this window the accumulation rate is flat rather than climbing, so the framework scores what is measured rather than what is projected beyond one window.
What to watch in the next 90 days
The first watch item is the next reserve release, which has no published date but on the observed 70-to-105-day spacing since Jun 19 2026 should land between Sep 2026 and Oct 2026; a firing materially larger than 21M LINK, or two firings inside one window, would push the forward reading above +3% and tip the inflation score from 1 to 0. The second is the split inside that release between the exchange destination and the staking-rewards multisig, since a larger multisig share reaches the open market more slowly. The third is the Chainlink Reserve accumulation rate, which has just printed its slowest fortnight of the window at 15,430 LINK a day; a sustained move higher is the only mechanism on this page that can compress the net figure. The fourth is Chainlink staking v0.3, still not live — a cap increase above the current 40.88M LINK community pool would be the first non-zero Buy #4 in this ledger. The fifth is any change to the published 7% per year release rate, or a first-ever burn of Reserve LINK; Chainlink has never burned a token, and either announcement would restructure this ledger entirely.
Summary
The MrNasdog Pressure Framework reads Chainlink at +2.57% net supply growth over the 90 days to Aug 24 2026 and +2.07% projected for the next 90 days — supply growing, and projected to keep growing, both readings inside the 1% to 3% band. The structural mechanism is unusual: LINK mints nothing and is capped at 1 billion forever, so every unit of dilution comes from a discretionary release out of a 251.9M LINK non-circulating reserve, most recently 21M LINK on Jun 19 2026, governed by a published rate of 7% of total supply per year that twelve months of on-chain reads reproduce exactly. The key risk is precisely that discretion — the rate is policy rather than protocol, the per-release size is known only after the transfer lands, and the overhang is large enough to fund years of releases. The offsetting force, the revenue-funded Chainlink Reserve, absorbed 1.8M LINK in the window and has never sent a token out, but it holds rather than burns, so at current rates it softens the dilution instead of ending it.
MrNasdog Pressure Framework analysis of Chainlink (LINK), Metric 1 — Inflation. Data + explanation only. Not financial advice. Updated Aug 24 2026.