A vesting cliff is the period after launch during which a bucket's tokens are fully locked and nothing releases at all. When the cliff date passes, vesting begins, and everything that accrued during the locked period usually becomes claimable at once. The cliff does not reduce how much supply eventually reaches the market. It decides which day the first tranche arrives and how large that tranche is.
A cliff concentrates rather than removes. On a four year vest, a one year cliff means twelve months of accrual lands in a single event, so the longer the cliff, the larger the first day.
Scroll to see the full diagram
A cliff is a gate, not a slope
During the cliff, the claimable balance is zero. Not small. Zero. Then on the cliff date the gate opens and the holder can claim whatever the schedule says has accrued, which on a standard design is the full amount that would have vested during the locked period.
That is the mechanic people miss when they treat the cliff as a safety feature. A four year vest with a one year cliff does not release one forty-eighth of the bucket on the cliff date. It releases twelve forty-eighths, a quarter of the bucket, in one event. Lengthen the cliff to eighteen months and that first event gets larger again.
Some designs break the accrual instead of banking it, so nothing accumulates during the cliff and monthly release starts fresh from the cliff date. That version genuinely reduces the first tranche, and it is worth specifying explicitly in the term sheet, because the two shapes are described with identical words and produce very different first days.
What the cliff is designed to produce
The cliff exists to make leaving expensive. A contributor who departs before the cliff date typically forfeits the entire allocation, since none of it has vested. That is a retention instrument borrowed almost unchanged from equity compensation, and it is the reason a one year cliff on team tokens is the convention rather than an arbitrary choice.
For investors the purpose is different. The cliff sets a floor under how long capital stays committed, which matters most in the window when a protocol has a token price and not yet a product. Sui's documentation is unusually direct about this: at mainnet launch there was a one year cliff period during which all initial investors were blocked from transferring their initial stake, and that period ended in May 2024.2
Neither purpose is served by a cliff that ends before the thing being incentivized has happened. If the protocol's first real usage milestone sits eighteen months out and the investor cliff ends at nine, the cliff has bought a commitment that expires before it is tested.
Two documented twelve month cliffs
Arbitrum's team, contributor and investor tokens vest over four years starting 16 March 2023, with the first unlock on 16 March 2024 and monthly unlocks after that.1 The twelve months between those two dates is the cliff, stated in the project's own documentation rather than inferred from chain data.
What arrived on the far side is the reason cliff length deserves modeling rather than defaulting. That first unlock released 1,111,750,000 ARB, roughly $1.24 billion at the prevailing price, equal to 87 percent of the token's then-circulating supply of 1.275 billion.4 The schedule performed exactly as written. The concentration was the design.
Sui took the same twelve month shape and a different disclosure posture, publishing the cliff and its end date while declining to publish bucket percentages at all.2 Between the two, a founder has two real reference points for what a one year cliff looks like when it breaks, which is two more than any benchmark table will give you.
The standard vesting contract has no cliff
This surprises people. OpenZeppelin's VestingWallet, the audited reference contract most vesting deployments are built on or modeled after, stores a beneficiary, a start timestamp and a duration, and releases strictly linearly against elapsed time.3 There is no cliff parameter in it.
Cliffs get implemented one of two ways as a result. Either the start timestamp is set in the future, which produces a hard gate with no accrual behind it, or the team deploys a purpose-built contract with an explicit cliff timestamp. Lockup platforms take the second route, exposing a cliff timestamp, a vesting duration and an optional immediate-unlock percentage as separate configurable fields.5
The practical consequence for diligence is that you cannot infer a cliff from the fact that a project uses a standard vesting contract. Read the constructor arguments. The difference between a future start timestamp and a cliff timestamp with accrual is the difference between a small first release and a quarter of the bucket.
Where cliffs actually go wrong
Rarely in the individual bucket. Almost always in the calendar. Investor, team and ecosystem grant cliffs get negotiated separately, each with a defensible length, and all three end inside the same sixty day window because twelve months from a single token generation event is twelve months for everybody. The result has its own name and its own page.
The fix is procedural and free. Put every cliff end date on one calendar before any term is signed, then deliberately stagger them. Moving one bucket's cliff by a quarter costs a conversation. Moving it after the agreement is executed costs a renegotiation with every holder in that bucket.
The second failure is quieter. Teams treat the cliff date as the moment risk begins, when the market has been pricing it since the terms were published. The date is public, the size is arithmetic, and the repricing tends to arrive ahead of the unlock rather than on it. Whether a specific token's cliff structure creates securities questions in your jurisdiction is fact-specific and belongs with counsel, and nothing here is a recommendation to buy, sell or hold anything.
Common questions
What does a one year cliff mean for tokens?
It means no tokens from that bucket are claimable for twelve months after the vesting start date. On the cliff date, vesting begins, and on most designs everything that accrued during those twelve months becomes claimable at once. On a four year vest that first event is a quarter of the bucket, which is why a longer cliff produces a larger first unlock rather than a gentler one.
What happens when a vesting cliff ends?
The claimable balance jumps from zero to whatever accrued during the locked period, and the holder can call the contract to claim it. Regular releases then continue on the schedule's cadence, usually monthly. Arbitrum's twelve month cliff ended on 16 March 2024 and released 1,111,750,000 ARB in that single event, an amount equal to 87 percent of circulating supply at the time.4
Is a longer vesting cliff always better?
No, and the assumption that it is causes real damage. A longer cliff banks more accrual behind the gate, so the first release grows with the cliff length. It also pushes every bucket's first unlock toward the same anniversary date if the cliffs are equal, which is how stacked cliffs form. Length is one variable, and the staggering across buckets usually matters more.
Do vesting cliffs apply to community and airdrop tokens?
Sometimes, and it is a design choice rather than a rule. Airdropped tokens are frequently fully unlocked on receipt, because the point is distribution and a locked airdrop converts poorly. Ecosystem and grant buckets often carry cliffs so the supply arrives alongside the programs it funds. What matters is that any cliff on these buckets goes on the same calendar as the insider ones.
See Token Allocation and Vesting Design for how this applies in practice.
Sources
- Token Supply
Arbitrum Foundation, 2023
Team, contributor and investor tokens unlock over four years from 16 March 2023, with the first unlock on 16 March 2024 and monthly unlocks thereafter. - Tokenomics on Sui
Sui Foundation and Mysten Labs, 2024
Documents the one year cliff at mainnet launch during which initial investors were blocked from transferring their initial stake, ending May 2024. - Contracts Finance API: VestingWallet
OpenZeppelin, 2023
Reference vesting contract storing beneficiary, start and duration, with strictly linear release against elapsed time and no cliff parameter. - Arbitrum Will Unlock $1.2B ARB in March 2024: Token Unlocks
Yahoo Finance, citing Token Unlocks, 2023
Reports the cliff unlock of 1,111,750,000 ARB on 16 March 2024, about $1.24 billion at $1.12, equal to 87 percent of the then-circulating supply. - Hedgey Community Documentation
Hedgey Finance, 2024
Vesting and lockup contracts exposing a cliff timestamp, a vesting duration and an optional immediate unlock percentage as separate configurable fields.
Last reviewed 2026-08
Know the terms but not sure how they apply to your project? That is what an engagement is for. We design, document, and stress-test the whole token economy inside the Tokenomics Data Room.
100+ projects advised. Complete tokenomics in 4 to 6 weeks.