Token Unlock and Vesting Benchmark: 18 Live Tokens Compared
An original benchmark of token unlock, vesting, and float structures across 18 live tokens. Every figure sourced. The pattern: unlock geometry, not allocation size, is the real risk.
A token unlock is the scheduled release of previously non-transferable tokens into circulating supply. A vesting schedule is the contract-level timetable that governs those releases: typically a cliff (a period with zero unlocks, most commonly 12 months) followed by linear distribution (equal monthly, bi-weekly, or daily tranches until the allocation is fully liquid). Together, the unlock schedule determines how a token's float grows from its launch percentage to full supply, and it is one of the few tokenomics decisions that is nearly impossible to change after launch.
Most writing on unlocks is anecdotal. This benchmark is not. We compiled the allocation, vesting, and unlock structures of 18 live tokens spanning L1s, L2s, DeFi, RWA, and infrastructure, from Uniswap's 2020 launch through Berachain and Story Protocol in 2025, using only figures verifiable in official documentation, foundation posts, and public unlock trackers. Every number traces to a cited source. Where a project does not publish a figure, we say so instead of guessing, and two candidate projects (io.net and Helium) were excluded entirely because their team and investor terms could not be verified from public sources.
The full machine-readable dataset is available as JSON alongside this page. Data captured July 25, 2026. Allocation and vesting terms are structural and stable; circulating-supply figures change daily and are flagged accordingly.
#Key findings
- The 12-month team cliff is a de facto standard. Of 12 projects that disclose a team cliff, 11 use exactly 12 months. The only outlier is Starknet, whose first team and investor unlock arrived roughly two months after TGE.
- Full team vesting clusters at 3 to 4 years from TGE (median 42 months across disclosed schedules). Ondo's 5-year core-team lock-up is the longest disclosed; 36-month totals (Celestia, EigenLayer, Jito, Jupiter, Berachain) are the shortest common pattern.
- Median TGE float is ~15% of total supply, but it is rising: the 2020-2023 launch cohort medians ~14% while the 2024-2026 cohort medians ~19%. This shift tracks the market's documented backlash against low-float, high-FDV listings.
- Insider share (team plus investors) medians 40% of supply. Four projects exceed 50% (Ethena and EigenLayer at 55%, Celestia 53.2%, Berachain 51.1%). Hyperliquid is the structural outlier at 23.8% with a zero-investor cap table.
- Every single-day unlock above 5% of supply in this dataset came from one design choice: team and investor cliffs expiring on the same calendar day. Ondo (~19.4%), Celestia (17.6%), Arbitrum (11.1%), and Berachain (~8.5%, derived) all share it. Projects that smoothed distribution (Starknet, Wormhole, Aptos) structurally cannot exceed ~2% in any single event.
#The benchmark table
Structural terms as published; percentages are of total supply at genesis. "n.p." means not published in any source we could verify. Full per-project records, notes, and source URLs are in the companion JSON.
| Project | Type | TGE | TGE float | Team % | Investor % | Team cliff | Team full vest (from TGE) |
|---|---|---|---|---|---|---|---|
| Uniswap (UNI) | DeFi DEX | Sep 2020 | 15.0% | 21.27% | 18.04% | n.p. (continuous) | 48 mo |
| Optimism (OP) | L2 | May 2022 | conflicting | 19.0% | 17.0% | n.p. | n.p. (events to 2029) |
| Aptos (APT) | L1 | Oct 2022 | 13.0% | 19.0% | 13.48% | 12 mo | 48 mo |
| Arbitrum (ARB) | L2 | Mar 2023 | 12.75% | 26.94% | 17.53% | 12 mo | 48 mo |
| Sui (SUI) | L1 | May 2023 | 5.0% | 20.0% | 14.0% | n.p. | n.p. |
| Sei (SEI) | L1 | Aug 2023 | 18% (disputed) | 20.0% | 20.0% | 12 mo | multi-year (to ~2029) |
| Celestia (TIA) | Modular L1 | Oct 2023 | ~26.7% | 17.64% | 35.57% | 12 mo | 36 mo |
| Jito (JTO) | DeFi / MEV | Dec 2023 | n.p. (airdrop 9%) | 24.5% | 16.2% | 12 mo | 36 mo |
| Ondo (ONDO) | RWA | Jan 2024 | ~14.3% | 33.0%* | 12.9% | 12 mo | 60 mo (5-yr team lock) |
| Jupiter (JUP) | DeFi aggregator | Jan 2024 | 13.5% | 20.0% | none disclosed | 12 mo | 36 mo |
| Starknet (STRK) | L2 | Feb 2024 | 7.28% (airdrop) | 20.04% | 18.17% | ~2 mo | ~37 mo |
| Ethena (ENA) | DeFi synthetic $ | Apr 2024 | ~19.6% | 30.0% | 25.0% | 12 mo | 48 mo |
| Wormhole (W) | Interoperability | Apr 2024 | 18.0% | 12.0% | 11.6%** | n.p. | n.p. (aligned to Oct 2028) |
| ZKsync (ZK) | L2 | Jun 2024 | ~35% | 16.14% | 17.19% | 12 mo | 48 mo |
| EigenLayer (EIGEN) | Restaking | Oct 2024 | ~12% | 25.5% | 29.5% | 12 mo | 36 mo |
| Hyperliquid (HYPE) | Perps DEX L1 | Nov 2024 | ~31% | 23.8% | 0% | 12 mo | multi-year (most 2027-28) |
| Berachain (BERA) | L1 | Feb 2025 | 21.5% | 16.8% | 34.3% | 12 mo | 36 mo |
| Story (IP) | IP L1 | Feb 2025 | 25.0% | 20.0% | 21.6% | n.p. | n.p. (investors 48 mo) |
* Ondo's "Protocol Development" bucket funds core contributors; core team members are additionally bound by a 5-year lock-up. ** Wormhole publishes no labeled investor bucket; "Strategic Network Participants" (11.6%) is the closest analog.
#The largest verified single-day cliff events
| Project | Date | Tokens unlocked | % of total supply | What unlocked |
|---|---|---|---|---|
| Ondo (ONDO) | Jan 18 cliffs, 2025-2029 | ~1.94B ONDO | ~19.4% | Annual cliff tranches across locked buckets |
| Celestia (TIA) | Oct 30, 2024 | ~175.6M TIA | ~17.6% of genesis | Year-1 cliff: ~58M core contributors + ~65M Series A/B + ~52M seed, same day (~80% of then-circulating supply) |
| Arbitrum (ARB) | Mar 16, 2024 | ~1.11B ARB | 11.1% | Year-1 cliff: 673.5M team/advisors + 438.25M investors, same day |
| Berachain (BERA) | Feb 6, 2026 | ~42.6M BERA | ~8.5% (derived) | Year-1 cliff: 1/6 of combined team + investor allocations |
| ZKsync (ZK) | Jun 2025 | ~756M ZK | 3.6% | Year-1 cliff for team + investors, then max 0.8%/month |
For contrast, the smoothed-schedule designs: Starknet's largest possible event is its 1.27% monthly tranche, Wormhole's W 2.0 runs bi-weekly unlocks of ~0.5%, and Aptos spreads its post-cliff releases into monthly tranches that peak around 2% (derived from its 3/48ths acceleration months). These schedules structurally cannot produce a double-digit supply day.
#What the cross-project data shows
#1. Team vesting has converged; float strategy has not
Team terms are the most standardized element in token design. The 12-month cliff appears in 11 of 12 disclosed schedules, and total vesting lands at 36 or 48 months in every disclosed case except Ondo's 60. Investor terms usually mirror team terms exactly (Arbitrum, Aptos, ZKsync, EigenLayer, Ethena, Jito, Berachain all use identical schedules for both groups). Celestia is the notable divergence in the other direction: its investors finish vesting at year 2 while its core contributors run to year 3.
TGE float, by contrast, spans a 7x range, from Sui's 5% to ZKsync's 35%. And the range is drifting upward. The 2020-2023 cohort in this dataset launched at a median ~14% float; the 2024-2026 cohort launched at ~19%, with Hyperliquid (31%), ZKsync (35%), and Story (25%) at the top. That shift followed public criticism of the low-float, high-FDV pattern: Binance Research's 2024 analysis found most major listings that year debuted with under 15% of supply circulating, many under 10%, and associated that structure with weaker post-listing performance and outsized sensitivity to later unlocks. The market read the same data and newer launches responded with bigger day-one floats.
#2. The mega-cliff is a design artifact, not a necessity
The most consequential pattern in the dataset: concentration risk on the calendar is chosen, not inherent. Ondo, Celestia, Arbitrum, and Berachain all pointed their team and investor cliffs at the same calendar day, and each produced a single-day supply event between 8.5% and 19.4% of total supply. Celestia's October 2024 cliff expanded circulating supply by roughly 80% in one day.
Independent research quantifies why this matters structurally. Keyrock's study of more than 16,000 unlock events across 40 tokens found that roughly 90% of unlocks created negative price pressure, that pressure typically begins about 30 days before the event as markets position for it, and that events releasing more than 5% of supply usually spark immediate volatility, with large unlocks averaging about 2.4x the price impact of small ones. The same study found recipient identity matters: team unlocks showed the worst average impact (drawdowns up to roughly 25% around the event), investor unlocks were more controlled (professional holders hedge via OTC and derivatives), and ecosystem unlocks were the one category with slightly positive average impact. These are historical observations about how markets have absorbed past unlock events, not forecasts about any token.
Every mega-cliff event in our table sits 2x to 4x above Keyrock's 5% volatility threshold. Meanwhile Starknet, Wormhole (after its W 2.0 conversion from annual cliffs to bi-weekly tranches), and Aptos demonstrate that the same total insider allocations can be released without ever crossing it. Wormhole's redesign is the cleanest natural experiment: the project explicitly replaced large annual cliffs with bi-weekly unlocks over 4.5 years, converting calendar risk into a steady drip.
#3. Insider share clusters at 40%, and the outliers are informative
Across projects with fully disclosed cap tables, combined team-plus-investor share medians 40% of total supply. The distribution's edges tell the story. At the high end, Ethena and EigenLayer (55% each), Celestia (53.2%), and Berachain (51.1%) allocated a majority of supply to insiders, which mechanically means the majority of supply arrives on a vesting timetable rather than through community distribution. At the low end, Hyperliquid allocated 23.8% to its team and 0% to investors, having never raised outside capital, and put roughly 70% of supply on the community side. Wormhole's 23.6% (12% team plus an 11.6% strategic-participants bucket) is the lowest conventional structure in the set.
#4. Disclosure quality is itself a benchmark dimension
Six of eighteen projects, including several with multi-billion dollar valuations, do not publish complete vesting terms for at least one core category. Optimism's docs describe multi-year linear unlocks without stating cliff lengths. Sui publishes an investor cliff but no team schedule. Wormhole retains unpublished "original agreements" for core contributors. Story states multi-year lockups without terms. Uniswap never published its cliff mechanics. Ondo documents a 5-year team lock but not a labeled team percentage (its Protocol Development bucket is the closest disclosed proxy). For anyone doing diligence on a token, the absence of a published per-category schedule is a finding in itself: it moves supply-side risk from calculable to unknown.
#A structural rubric: healthier vs riskier unlock design
Derived from the patterns above and the cited unlock research. This is a mechanism-design checklist, not investment advice; whether any given structure fits a specific project depends on its business model, treasury needs, and distribution goals.
| Dimension | Healthier pattern (per this dataset) | Riskier pattern (per this dataset) |
|---|---|---|
| Team cliff | 12 months or longer (the market norm: 11 of 12 disclosed) | Under 12 months, or undisclosed |
| Team full vest | 36-48 months from TGE; longer signals longer alignment | Under 36 months |
| Investor terms | Equal to or longer than team terms | Investors vesting materially faster than the team |
| TGE float | Roughly 15-30% of supply, enough for price discovery | Under 10% float with a large FDV overhang, or a float built on short-dated locks |
| Unlock geometry | Monthly, bi-weekly, or daily tranches; no single event above ~5% of supply | Team and investor cliffs expiring the same day (produced 8.5-19.4% single-day events in this dataset) |
| Insider share | At or below the ~40% median | Above 50% of total supply |
| Disclosure | Full per-category schedule in official docs | Terms available only via third-party trackers, or not at all |
The single highest-leverage fix visible in this data is unlock geometry. A project can keep identical allocations and identical total vesting length, and still remove its largest structural risk event, by de-synchronizing team and investor cliffs or converting the cliff release into tranches. Wormhole's W 2.0 conversion and Starknet's capped monthly tranches are live examples of the pattern.
#Methodology
- Scope: 18 live tokens selected for structural diversity (L1, L2, DeFi, RWA, interoperability) and launch-era diversity (2020 through 2025). Two candidates were excluded for insufficient public disclosure: io.net (team and investor vesting terms not verifiable) and Helium (emission-based distribution with no published insider vesting schedule, structurally not comparable).
- Sources: Official project docs and foundation posts were preferred for allocation and vesting terms; public trackers (Tokenomist/TokenUnlocks, DropsTab, CryptoRank, CoinGecko, CoinMarketCap) and exchange research (Binance Research) for unlock events and supply snapshots. Every per-project figure is source-linked in the companion JSON.
- Conflicts: Where reputable sources disagree (common for circulating supply, occasional for TGE float), we report the conflict and the range rather than picking a number. Figures we computed from official vesting rules rather than found directly stated (Berachain's cliff size, Aptos's accelerated-month rate) are labeled derived.
- Date sensitivity: Circulating-supply percentages are snapshots as of July 25, 2026 and change continuously. Allocation percentages and vesting terms are fixed at launch (exceptions noted: Jupiter's 3B token burn, Wormhole's W 2.0 schedule conversion, Aptos's supply-cap governance vote).
- What this is not: This benchmark describes supply structures. It makes no forward-looking claims about any token's price and recommends no transaction in any asset.
#Sources
Primary documentation and per-project sources:
- Arbitrum Foundation: Token Supply; ARB March 2024 unlock coverage
- Binance Research: Optimism
- Starknet docs: STRK token
- ZK Nation docs: ZK token
- Celestia docs: Staking, governance, and supply; The Block: Celestia October 2024 unlock; CoinDesk: TIA unlock breakdown
- Aptos: Tokenomics Overview
- Sui: Token Schedule; Sui docs: Tokenomics; Sui blog: Understanding SUI token supply
- Sei blog: Future-Proofed Tokenomics
- Jupiter tokenomics coverage: CoinDesk JUP; Blockworks: JUP supply analytics
- Ethena docs: ENA Tokenomics
- EigenLayer EIGEN launch allocation (Binance summary); CoinDesk: EIGEN trading debut
- Jito governance docs: The JTO token
- Hyperliquid tokenomics: TokenRadar; Collective Shift: HYPE; Yahoo Finance: HYPE launch
- Wormhole: W Tokenomics; Wormhole: W Token 2.0 upgrade
- Berachain docs: BERA token; Binance Research: Berachain
- Story Protocol tokenomics: Tokenomist; Gate Learn: Story IP tokenomics guide
- Ondo Foundation docs: ONDO token; DropsTab: ONDO vesting
- Uniswap: Introducing UNI
Cross-project research cited in the analysis:
- Keyrock: From Locked to Liquidity, What 16,000+ Token Unlocks Teach Us
- Binance Research: low float / high FDV listing dynamics (2024)
Supply snapshots additionally reference CoinGecko, CoinMarketCap, Tokenomist, DropsTab, and CryptoRank per-project pages as of July 25, 2026; complete URL lists per project are in token-unlock-benchmark-data.json.