CoinYQ Dossier

Tezos wrote an amendment clause, then let the clause rewrite the economy

The 2014 Tezos paper treated a blockchain less like a finished machine than a charter with a procedure for replacing its own rules. Mainnet later used that procedure to change gas limits, consensus, issuance, staking and rollup infrastructure. The result is not governance without leaders: code still comes from identifiable teams, capital gathers behind bakers, and Etherlink has its own sequencer. Tezos instead makes those transitions legible as votes, activations and distinct control surfaces.

A 2014 constitution became a recommended 763.3 million-tez genesis

The September 2014 whitepaper proposed a functional protocol module that could be replaced by stakeholder-approved code, including changes to the amendment process itself. It was an attempt to keep technical evolution inside one ledger rather than force every disagreement into a new chain.

The July 2017 fundraiser later mapped contributions into a recommended genesis allocation. Foundation statistics report 607,489,040.89 tez for participants, 3,156,502.85 for early backers and contractors, and two allocations of 76,330,692.97 for the Tezos Foundation and Dynamic Ledger Solutions. Together they made 763,306,929.69 tez, with the two institutional allocations—20% in total—scheduled to vest in monthly installments over four years. These are the Foundation’s recommended allocation statistics, which it says could change slightly, not an independent replay of final genesis balances. Betanet opened on June 30, 2018; holders could accept that allocation or run the open-source protocol with another genesis.

Athens proved that a protocol hash could win an election

Athens became the first autonomous amendment in May 2019. The mechanism now runs through five periods: delegates submit and rank protocol hashes, vote in Exploration, use Cooldown to test, vote again in Promotion, then prepare during Adoption before automatic activation. The ballots require a dynamic quorum and more than 80% Yea among decisive votes. Pass ballots count toward quorum but not the Yea-to-Yea-plus-Nay ratio.

Only delegates cast those ballots. A holder who stakes or delegates assigns voting weight to a baker; holding liquid XTZ without delegate status is not a direct vote. Developers can publish code and campaign for it, but a proposal hash still needs baker voting power and nodes able to run the activated result.

Paris made inflation answer to the amount actually staked

XTZ pays fees and supplies the security deposit behind baking. It has no fixed maximum. Paris activated Adaptive Issuance in 2024: instead of treating annual issuance as a permanent percentage, the protocol computes a reward budget from total supply and the share actively staked, nudging that share toward a 50% target within protocol bounds.

That change also sharpened the difference between staking and delegation. A baker needs at least 6,000 tez of baking power. Staked tez is locked, counts fully, earns protocol-distributed rewards and can be slashed with the baker. Delegated tez remains spendable, counts one-third for baking, carries no slashing loss and sends rewards first to the baker, who is not protocol-bound to pass them on.

Both forms count fully toward the baker's governance power. A delegator keeps custody and liquidity but hands a political input to the baker. A staker keeps ownership in the account but accepts an unstaking delay and the baker's consensus risk. Neither choice grants ownership of a baker business or the Foundation treasury.

The Foundation can fund the pen but cannot cast the final code by itself

The Tezos Foundation is a Swiss foundation whose Council and Executive Committee manage its assets, grants and strategy. It explicitly says it is not Tezos and does not operate the network. Its grant program finances core and infrastructure work by independent teams, including Nomadic Labs, Trilitech and Functori.

That separation is real but not weightless. Funders choose which engineering, security and adoption work receives capital; core teams maintain Octez and draft upgrade packages; bakers decide which hash advances. Tezos has no layer-1 administrator key that silently patches the economic protocol, yet agenda setting, client concentration and delegated voting power remain forms of control outside the final ballot.

Ushuaia split the layer-1 amendment clock from Etherlink’s kernel clock

Tallinn activated a 6-second layer-1 block time on January 24, 2026. Ushuaia followed on June 30 with 10 MB/s Data Availability Layer bandwidth, dynamic DAL confirmation and a route for Etherlink PVM features to activate through rollup governance. Its enshrined liquid staking and quantum-resistant account experiments stayed behind testnet-only flags.

Etherlink is not a synonym for layer 1. Its EVM transactions normally pass through one sequencer account, with a delayed inbox as the censorship-resistance fallback. Separate contracts govern slow kernel releases, fast security releases and the sequencer operator, using Tezos baker voting power.

That distinction mattered in August 2026. Bakers were asked to reject an Etherlink 7.0 slow-governance candidate after testing found a vulnerability; a patched kernel then activated through fast governance on August 21. The live Michelson interface is evidence of a shipped Tezos X component. It does not turn every previewnet or roadmap item into a Mainnet right for XTZ holders.

How the project changed

  1. 2014-09-02
    The self-amending ledger is published

    The whitepaper describes stakeholder-approved meta-upgrades and proof-of-stake smart contracts.

  2. 2017-07
    The fundraiser defines the recommended genesis

    Bitcoin and ether contributions map to 763,306,929.69 initial tez, including two 10% institutional allocations.

  3. 2018-06-30
    Tezos betanet opens

    The primary network begins with fundraiser allocations and an open-source protocol others could fork.

  4. 2019-05
    Athens activates autonomously

    The first approved amendment changes gas and the baking threshold through the chain’s own election.

  5. 2024-06-05
    Paris activates Adaptive Issuance

    Staking and a supply-ratio-driven reward budget replace the earlier fixed-reward economics.

  6. 2026-01-24
    Tallinn shortens layer-1 blocks to 6 seconds

    The twentieth amendment lowers latency while preserving the same amendment route.

  7. 2026-06-30
    Ushuaia expands the DAL

    10 MB/s data availability and rollup-aligned PVM activation go live; experimental liquid staking remains testnet-only.

  8. 2026-08-21
    Patched Etherlink 7.0 activates

    After rejection of a vulnerable candidate, fast governance ships the Michelson interface and native atomic composability.

Evidence and primary sources

Last evidence review: 2026-09-04

What is Tezos?

Tezos is a self-amending proof-of-stake blockchain whose native tez (XTZ) pays fees, secures baking and carries delegate voting power. Its protocol can replace itself after a five-period baker vote; staking and delegation assign different reward, lock and slashing rights, while Etherlink runs as a separately governed EVM Smart Rollup above Tezos layer 1.

What problem does Tezos solve?

Tezos began with a governance problem rather than a throughput slogan: how can a blockchain change its own rules without treating every upgrade as an improvised social fork? Its answer embeds proposal hashes, quorum, supermajority votes and automatic activation in the protocol. That machinery coordinates upgrades, but it does not erase politics. Core teams still write code, bakers aggregate other holders' voting weight, node operators must install compatible software, and layer-2 systems such as Etherlink maintain their own kernels and operators.

How does Tezos work?

On layer 1, bakers stake tez, run Octez and DAL infrastructure, propose blocks and attest to others. Users may stake with a baker, locking tez and sharing protocol rewards and slashing exposure, or merely delegate liquid tez, which gives the baker voting and partial baking power while leaving reward payment to the baker. Protocol amendments move through Proposal, Exploration, Cooldown, Promotion and Adoption periods, with more than 80% Yea among Yea plus Nay in the two ballots; Pass counts toward quorum but not that ratio. Adaptive Issuance changes new-coin rewards with the staked-supply ratio. Etherlink posts rollup data and commitments to Tezos but has separate slow and fast kernel votes plus a vote over the single sequencer operator.

Key facts

  • The September 2, 2014 whitepaper described a generic crypto-ledger whose protocol and even amendment procedure could amend themselves.
  • The Foundation’s recommended genesis statistics total 763,306,929.69 tez: 607,489,040.89 for fundraiser participants, 3,156,502.85 for early backers and contractors, and 76,330,692.97 each for the Tezos Foundation and Dynamic Ledger Solutions. Those two institutional allocations together account for 20%, scheduled to vest in monthly installments over four years; these are published allocation statistics, not an independently replayed genesis balance.
  • Tezos Mainnet opened as betanet on June 30, 2018. The current Octez documentation identifies Ushuaia as the active protocol reviewed on September 4, 2026.
  • Only registered delegates vote on layer-1 protocol amendments. Their voting power includes their staked tez and tez delegated to them; ordinary XTZ ownership does not itself submit a ballot.
  • Each amendment stage lasts 14 cycles, about 14 days. Exploration and Promotion require dynamic quorum and more than 80% Yea among Yea plus Nay votes.
  • A baker needs at least 6,000 tez of baking power. Staked tez counts 100% toward baking power and can be slashed; delegated liquid tez counts 33% toward baking power and is not slashed.
  • The protocol pays staking rewards. The reviewed staking guide describes withdrawal maturity as the remainder of the request cycle plus three full cycles, up to about four days, followed by finalization. Ushuaia’s reference sets UNSTAKE_FINALIZATION_DELAY to three cycles; the separate two-cycle consensus-rights lookahead is not a withdrawal deadline. Delegation rewards first go to the baker, whose onward payment is not guaranteed by the protocol.
  • XTZ has no fixed maximum supply. Adaptive Issuance, active since the Paris era, adjusts the reward budget using total supply and the staked-supply ratio around a 50% target instead of preserving a fixed annual inflation rate.
  • The Tezos Foundation manages its own treasury and grants but says it neither manages nor operates Tezos. Funded teams such as Nomadic Labs, Trilitech and Functori can build clients and proposals; bakers decide protocol activation.
  • Ushuaia activated on June 30, 2026 with 10 MB/s DAL bandwidth and rollup-aligned PVM upgrades. Enshrined liquid staking and quantum-resistant keys in that release were testnet-only feature flags, not Mainnet products.
  • Etherlink is an EVM-compatible Smart Rollup with one sequencer account and a delayed layer-1 path. Tezos bakers separately vote on slow/fast kernel upgrades and on the sequencer operator.
  • Etherlink 7.0 activated on August 21, 2026 after a vulnerable slow-governance candidate was rejected and a patched kernel used fast governance. Its Michelson interface is live; broader 'Tezos X' claims should be checked feature by feature.

Official links

Categories

Related coins

Frequently asked questions

Does every XTZ holder vote directly on Tezos upgrades?

No. Layer-1 ballots are signed by registered delegates. Staked and delegated tez contribute to a delegate’s voting power, so choosing a baker also assigns that baker the voting weight of the account unless the holder becomes a delegate.

Can Tezos really upgrade without any hard fork?

An approved protocol hash activates automatically after the five-period amendment process, avoiding a conventional coordinated fork for participating Mainnet nodes. Operators still install compatible software, and people remain free to reject the social consensus or launch another network.

What is the practical difference between staking and delegation?

Staking locks tez, earns protocol-paid rewards, contributes full baking power and shares the baker’s slashing exposure. Delegation leaves tez liquid and unslashed, contributes one-third baking power, and relies on the baker for any reward payment. Both contribute full voting power to the baker.

Is XTZ supply capped?

No fixed maximum is documented. Protocol rewards create tez, and Adaptive Issuance changes the nominal issuance budget according to the staked ratio and protocol bounds. A current inflation figure is therefore a live network value, not a permanent token term.

What can be slashed?

A baker’s stake and the tez users actively stake with that baker can be penalized for offenses such as double baking or double attesting. Merely delegated, liquid tez is not exposed to slashing, although the delegator may miss rewards if the baker performs poorly or does not pay.

Does XTZ represent ownership of the Tezos Foundation treasury?

No such property or redemption right is documented. XTZ is the network asset for fees, staking, delegation and governance weight. The Swiss Foundation controls its own assets and grant decisions under its institutional governance.

Is Etherlink just the newest Tezos protocol?

No. Etherlink is an EVM Smart Rollup anchored to Tezos. It has its own kernel, a single elected sequencer operator and separate governance contracts, even though Tezos bakers provide the voting weight and layer 1 supplies data availability and dispute security.

Are all Tezos X roadmap features live?

No. Ushuaia made major DAL and rollup changes live, and Etherlink 7.0 added its Michelson interface. The same Ushuaia package explicitly kept enshrined liquid staking and quantum-resistant keys behind testnet-only flags, so roadmap labels cannot substitute for deployment evidence.

External trackers

Choose a tracking site for Tezos: