SEI Inflation Analysis · August 2026 · Supply growing, projected to keep growing

Sei adds 311.0M SEI to the market every 90 days — 266.7M from monthly team and private-sale vesting cliffs and 44.3M from a token release schedule written into the chain — and removes nothing, because Sei has no burn, no buyback and no Foundation bid. That is +4.62% net over the last 90 days and +4.57% projected over the next. Our supply monitor reads −0.03% for the same window, a gap of 4.65 percentage points that ships a ⚠ monitor gap chip, because the circulating figure the market quotes for SEI has not moved off 6.73B since Mar 15 2026. Sei's 10B cap is real; it is also five years away from binding.

The verdict, in one paragraph

For the 90-day window ending Aug 18 2026, the MrNasdog Pressure Framework reads SEI at +4.62% net, with +4.57% projected over the next 90 days. Our supply monitor reads −0.03% for the trailing window, a gap of 4.65 percentage points — far outside the framework's 0.5-point tolerance, so a ⚠ monitor gap chip is attached to the SEI overview page. The gap is not a disagreement about Sei's mechanics; it is a stalled denominator. The classified circulating figure has held exactly 6,733,333,333 SEI for 156 days, and exactly 6,127,777,777 for the 73 days before that, so a monitor built on it can only ever read rounding noise around zero. A second classifier publishes 7.46B on the same day — a 727M divergence that is roughly the flow this analysis measures, accumulated while the first field sat still. SEI is a capped token on an uncapped release calendar: nothing about the 10 billion ceiling slows the rate at which SEI reaches the market before it.

Sell pressure: where new SEI comes from

Sell #2 — vesting unlocks — is the larger force, at 266.7M SEI in the last 90 days and the same 266.7M SEI projected. Sei's genesis split 10 billion SEI five ways: an Ecosystem Reserve at 48%, Team at 20%, Private Sale Investors at 20%, the Sei Foundation at 9% and a Binance Launchpool tranche at 3%. The Team and Private Sale allocations, 2B SEI each, cleared a one-year cliff in August 2024 and have released on the 15th of every month since — 33.3M SEI from the team tranche on a five-year line and 55.6M SEI from the investor tranche on a three-year line, 88.9M SEI together. Three of those cliffs landed inside this window: Jun 15 2026, Jul 15 2026 and Aug 15 2026. Three more land in the next.

Sell #1 — protocol inflation — adds 44.3M SEI, and it works unlike almost any other proof-of-stake chain in the catalogue. Sei's mint module does not strike an inflation rate against staked supply; it reads a token release schedule encoded in the chain and hands out a fixed annual quantum split evenly across the days of each period. Those quanta step down every year: 300M SEI for the first year from Aug 15 2023, 225M, then 180M, then 165M for the year beginning Aug 15 2026, falling to 75M by 2032 — 1.5B SEI in total, which with the 8.5B minted at genesis is exactly the 10 billion cap. The step from 180M to 165M landed three days before this build, inside the window, so the trailing figure blends two rates (494,505 SEI a day for 86 days, 453,297 SEI a day for four) while the forward projection uses only the live one: 40.8M SEI. Total SEI supply now stands at 9,206,812,155, which matches the cumulative schedule to within 1,032 SEI — proof that nothing has been minted off-plan and nothing has been destroyed.

The other two sell rows are zero. Sell #3 — Foundation and unscheduled unlocks — books zero for a double-counting reason covered below rather than an absence of holdings. Sell #4 — long-term locked or bankruptcy — is zero because no estate, trustee or court-administered pool holds SEI at all.

Buy pressure: where new SEI goes

Nowhere. All four buy rows read zero, and this is the fact that decides the SEI page. Buy #1 — programmatic buyback — is zero because Sei operates no repurchase contract, no burn auction and no published programme that spends protocol revenue on SEI; there is consequently no accumulation wallet behind the row and no destination question to answer. Buy #2 — protocol fee burn — is zero because Sei deliberately did not copy Ethereum's base-fee burn: gas paid on Sei is pooled and paid out to validators and delegators, so a busy quarter raises staking income rather than shrinking supply. Both routes to destruction were checked. Supply itself has never fallen, which rules out an event-driven burn; and the two dead addresses on Sei's EVM layer hold 19 SEI and 1,386 SEI between them, which rules out the transfer-to-nowhere route.

Buy #3 — Foundation buy — is zero: the Sei Foundation runs no open-market purchase programme with a published amount or a set schedule, and no corporate treasury discloses SEI purchases on a timetable. Buy #4 — new long-term lock — is zero: 4.19B SEI is bonded to validators, but bonding carries a 21-day unbonding wait, which is a delay rather than a lock, and no new escrow or vault was deployed inside the window. A ledger with 311.0M SEI on the sell side and a hard zero on the buy side has no offsetting mechanism to discuss — the net is simply the gross.

Foundation and overhang

Sei's overhang is large and mostly on a published calendar. The Ecosystem Reserve of 4.8B SEI is the biggest team-controlled pool, and the part of it that actually reaches the market is the chain's own release schedule — the 1.5B SEI already booked in Sell #1. Charging a modelled reserve curve on top of that on-chain release would count the same tokens twice, so Sell #3 stays at zero and the reserve is tracked as scope instead. What remains inside it is discretionary grant and ecosystem spend from wallets the Sei Foundation does not label; no outflow was observable at this refresh. The Foundation's own 900M SEI tranche finished releasing in 2023 and the Binance Launchpool's 300M SEI went out at launch, so neither holds anything back.

The remaining overhang is the unreleased half of the Team and Private Sale tranches — roughly 1.2B SEI still to come from the team line through Aug 2029 and 667M SEI from the investor line through Aug 2027 — plus the 793M SEIthe release schedule has not yet created. None of it sits in a readable lock contract: Sei's genesis file contains only plain accounts, so these are contractual holds in unlabelled wallets, refreshed by a walk of the published schedule and the project's own disclosure. If any of those balances falls between refreshes, the outflow enters Sell #3 at the next refresh.

How SEI compares to other capped proof-of-stake chains

SEI belongs to an unusual class: hard-capped proof-of-stake L1s that fund staking from a pre-minted reserve rather than from open-ended issuance. That design is often marketed as the best of both worlds — a fixed 10 billion ceiling like Bitcoin's 21 million, with none of the perpetual dilution a Cosmos mint module imposes. The Pressure Framework reads it differently. A cap constrains where supply ends, not how fast it arrives, and Sei is arriving quickly: 9.21B of the 10B cap already exists, while the market counts only 6.73B as circulating. The distance between those two numbers is the whole story, and the vesting calendar closes it at roughly 89M SEI a month.

Against an uncapped Cosmos chain running a bonding-ratio mint, SEI looks better on paper and worse in practice. An uncapped chain typically issues 5–8% a year against its full supply and pays it to stakers, who are long by design; Sei issues 165M SEI a year on a schedule that only falls, which is genuinely more disciplined — but then adds 1.07B SEI a year of cliff-driven distribution to team and investor wallets, which the uncapped chain does not have because its genesis vesting has usually expired. Against halving-model proof-of-work chains, the contrast is sharper still: a halving cuts issuance by 50% at a stroke and no vote can raise it, whereas Sei's step-downs are gentle (180M to 165M is an 8% cut) and its dominant supply force is a calendar, not a subsidy.

The most instructive comparison is with chains that pair issuance to a fee burn or a revenue buyback. Those designs give supply a counterweight that scales with usage: a busy quarter can flip an EIP-1559 chain deflationary, and an exchange-token buyback converts revenue directly into destruction. Sei has neither, by choice — gas revenue funds staking rewards instead — so Sei's network activity, however fast the Giga upgrade makes it, cannot touch the supply side of this page. Sei is also, usefully, a single-ledger token despite running a Cosmos layer and an EVM layer: both spend the same balance, there is no ERC-20 origin contract and no bridge lockbox, so one supply read is the whole picture — a rarity worth stating plainly.

What to watch in the next 90 days

Three vesting cliffs fall inside the window — Sep 15 2026, Oct 15 2026 and Nov 15 2026 — each releasing about 88.9M SEI to team and private-sale wallets, and together they are 87% of this reading's sell side. Second, watch the release schedule's next step: the current period runs to Aug 14 2027 at 165M SEI a year before dropping to 135M, so nothing changes on the emission line inside this window and any change would have to come from governance. Third, the Private Sale tranche exhausts in Aug 2027, which will cut the monthly cliff from 88.9M to 33.3M SEI — the single largest scheduled improvement in Sei's supply picture, and the first date at which this page could plausibly read under 2%. Fourth, any Sei governance proposal introducing a fee burn or a buyback would be the first buy-side mechanism the token has ever had. Finally, the classified circulating figure itself: if it is ever unfrozen from 6.73B, the monitor gap on this page closes without a single thing changing on the Sei chain.

Summary

The MrNasdog Pressure Framework reads SEI as clearly and persistently inflationary: 311.0M SEI reaching the market per 90 days against zero removed, a net of +4.62% realised and +4.57% projected. The structural mechanism is two calendars running at once — a chain-encoded token release schedule paying 165M SEI a year to stakers, and monthly Team and Private Sale cliffs releasing 88.9M SEI on the 15th — with no burn, no buyback and no Foundation bid anywhere in the design to offset either. The key risk is that the cap invites the wrong conclusion: Sei's 10B SEI ceiling is real and 9.21B of it already exists, but the tradable float is only 6.73B, and the gap between those two numbers is what the vesting calendar spends the next three years closing. The ceiling binds eventually; nothing binds before it.

MrNasdog Pressure Framework analysis of SEI, Metric 1 — Inflation. Data + explanation only. Not financial advice. Updated Aug 18 2026.