Every number on this site is either read from a contract on this request or stated with its source. This page is where you check that. It includes the things that do not flatter us.
This balances to the token. Add the exclusion list below to the circulating figure on BNB Chain and you get the contract’s total supply exactly. The cross-chain headline above is larger because TRON and Polygon are separate contracts with their own supply It is not part of this identity, and adding it to the exclusion list would compare two different scopes.
How circulating supply is computed, address by address.
XPHX is three separate contracts, so circulating supply is the sum across BNB Chain, TRON and Polygon. Each chain is totalSupply() minus a published exclusion list, and every excluded address is queried live on the request that renders the page, so you can total them on a block explorer and arrive at the same figure.
The methodology page also records the holdings we deliberately count against our own interest, the deviation from our October 2024 CoinMarketCap submission, and a 12,400,001 SZNX loss, read on this request, that is permanently unrecoverable.
Open the full methodologyThe 11M mining allocation. Not total supply.
11,000,000 XPHX, 33.33% of the cap, is set aside to pay mining and staking rewards. Nothing outside that allocation is ever used for them. The schedule runs eight program years from 2021-10-01, October to September.
Some historical rewards vest over time rather than landing in full when earned, which is why released trails granted. Granted and released are therefore different quantities, and the difference is an obligation we fund rather than a shortfall.
| Channel | Since | Custody | Granted | Released |
|---|---|---|---|---|
| PhoenixCo CeFi website staking | 2021 | custodial | 3,745,666measured · split by year below | 2,176,975 |
| Third-party DeFi poolsBabyDogeSwap · SafeSwap · DODO · MelegaSwap | 2023 | non-custodial | 5,539,334derived · not splittable by year | ≤ 4,681,5065,539,334 − 857,828 still in 16 contracts |
| PhoenixCo DeFi Hub | 2026 | non-custodial |
Every pool holding XPHX, read live, with its address.
Reading pool balances…
Five years and twelve pools before this app existed.
The pools in this app are not a new product. PhoenixCo has run staking since October 2021, 4 years and 12 pools on the original site, granting 3,745,666 XPHX to stakers. The DeFi Hub is that programme continued on chain, where the accounting can be checked by anyone instead of taken on trust.
A database snapshot taken 2026-08-08, not a live read. These come from the old site's own books, which are not public. Pools 11 and 12 are still running, so both granted and released keep moving after the export. The DeFi Hub pools in this app are the opposite: read from chain on every request, so nothing about them is a snapshot.
A Compound v2 fork, deployed by us, holding real supply and real debt.
24 markets, each an independent pool with its own interest-rate curve, its own collateral factor and its own caps. Supplying earns interest paid by borrowers; borrowing draws against collateral you have separately enabled. Nothing is pooled across markets except the risk engine that prices them.
The Unitroller is the contract that governs. It holds the market list, the collateral factors and the oracle address, and it is what a liquidation is settled against. Every address this interface reads is checked against it before a deploy. If the two disagree, the deploy fails rather than showing risk numbers from a contract the protocol does not consult.
Tiers change limits, never access. Locking XPHX raises the fraction of your collateral you may borrow against, read live from the risk engine. It does not gate any market, and it does not change what anybody else can do.
Audit status. The lending engine is a fork of audited Compound V2; our additions have been reviewed internally and not yet by a third party. The terms page sets out what a borrower is exposed to before they borrow.
Every deployment, on every chain. Verify before you send anything.
Two audited bases, what we changed, which key can do what, and where the gaps are.
The staking pools are Synthetix StakingRewards with eighteen documented changed lines (counted 1 August 2026), every one forced by the Solidity 0.8 migration, and the reward maths pinned byte for byte by a test. The lending markets are Compound V2: the same Unitroller, Comptroller and cToken design behind the largest lending protocols, reviewed more times than almost any code in DeFi. Neither base is ours. Everything below the line is.
Something here wrong or missing? It is a bug, and we want to know.
| 28,875100,000 − 71,125 still in 2 contracts |
Why one figure carries a ≤. Third-party DeFi pools pays out through venues whose reward contracts are not all recorded here, so only 857,828 of unpaid rewards can be subtracted. Whatever still sits in SafeSwap and MelegaSwap is counted as released when it has not been, so the true figure is at most the one shown. One of those programmes has no recorded size, so no floor can be put under it either.
Sum: 9,385,000 granted of 11,000,000, 85.32%, leaving 1,615,000 for roughly three more years. Stated as at 8 August 2026.
Above the year-five plan, on a comparison that is not like for like. Planned cumulative RELEASE by the end of programme year five was 8,360,000; actual granted is 9,385,000, which is 1,025,000 more, about 12.3% above it. Rewards are paid per block. BNB Chain's block time decreased over the period, so the same per-block rate delivered more tokens per calendar day than the schedule assumed.
| Program year | Planned | Cumulative | Granted | Released |
|---|---|---|---|---|
| 2021-22 | 2,200,000 | 2,200,000 | pending | pending |
| 2022-23 | 1,870,000 | 4,070,000 | pending | pending |
| 2023-24 | 1,650,000 | 5,720,000 | pending | pending |
| 2024-25 | 1,430,000 | 7,150,000 | pending | pending |
| 2025-26 | 1,210,000 | 8,360,000 | pending | pending |
| 2026-27 | 990,000 | 9,350,000 | pending | pending |
| 2027-28 | 880,000 | 10,230,000 | pending | pending |
| 2028-29 | 770,000 | 11,000,000 | pending | pending |
| Total | 11,000,000 | 11,000,000 | pending | pending |
One channel splits by year; the other two cannot. The CeFi channel does, and it is derived from the pools themselves rather than apportioned: every pool has a measured total, a start date and a term, and rewards accrue per second, so accruing each pool linearly across its own term is how the product works. The twelve sum to 3,745,666 exactly. The third-party channel is a derived residual with no per-pool ledger behind it, so splitting it would invent a shape for a number that has none. The DeFi Hub is an allocation with a start date, not an accrued grant.
| Year | Granted | State |
|---|---|---|
| 2021 | 652,714 | complete |
| 2022 | 1,592,752 | complete |
| 2023 | 475,022 | complete |
| 2024 | 555,949 | complete |
| 2025 | 327,935 | complete |
| 2026 | 112,888 | partial |
| 2027 | 25,502 | scheduled |
| 2028 | 2,904 | scheduled |
Rewards on the old website vested rather than paying out on the day they were earned, so the amount granted and the amount actually in wallets are different numbers. Only 58% of what was granted has reached a wallet.
Granted = in wallets + still vesting (2,176,975 + 1,568,692 = 3,745,667, against 3,745,666 granted, a one-token rounding difference). The 27,031 shown as already unlocked is part of the vesting figure, not a third bucket on top of it.
| Pool | Allocation | Granted | In wallets | Started | Status |
|---|---|---|---|---|---|
| Pool 1 | 1,040,000 | 944,739 | 944,739100% | 2021-10-01150d | closed |
| Pool 2 | 330,000 | 290,708 | 282,56797% | 2021-10-01365d | closed |
| Pool 3 | 470,000 | 474,913 | 165,28735% | 2022-02-03190d | closed |
| Pool 4 | 525,000 | 525,098 | 223,67843% | 2022-03-03365d | closed |
| Pool 5 | 130,000 | 133,304 | 77,97858% | 2022-09-03365d | closed |
| Pool 6 | 330,000 | 328,123 | 156,33248% | 2022-09-03730d | closed |
| Pool 7 | 63,000 | 63,003 | 38,98462% | 2023-03-21365d | closed |
| Pool 8 | 707,000 | 702,631 | 225,20632% | 2023-10-17633d | closed |
| Pool 9 | 70,000 | 70,003 | 13,99820% | 2024-09-13533d | closed |
| Pool 10 | 100,000 | 83,504 | 38,17146% | 2024-09-13800d | closed |
| Pool 11 | 180,000 | 111,809 | 9,7159% | 2025-07-19621d | running |
| Pool 12 | 165,000 | 17,831 | 3192% | 2026-05-25700d | running |
These figures are stated from the programme's own books, not computed on chain. The CeFi product is custodial, so its ledger is not public the way the pool contracts in this app are. That is exactly why the programme moved on chain.
Staking pool addresses are listed on each pool's page and in a registry committed to our public repository rather than held in a database, so the set of pools this site will read is fixed at build time and reviewable.
These five contracts were deployed on BNB Chain and then replaced before they were ever funded. They hold no reward budget, report a reward rate of zero, and are not part of the live platform. They are listed so nobody mistakes one for an active pool, because staking into any of them would lock your tokens against rewards that will never be paid. The live pool addresses are the ones on this site's pool pages.
pauseGuardian() returns the zero address: no separate guardian has been set, so the admin key is the only one that can pause, and it can also unpause, change collateral factors, add markets and set oracles. That admin is an externally owned account rather than a multisig or a timelock. Both are on the list, and until they are done, the honest description of this protocol’s governance is one key.Every address in the contracts section links to BscScan. The staking pools, the lending markets and the tier locker are verified on BscScan, with source published and matched to the deployed bytecode. Verification of the sXPHX vault is in progress. Verification lets anyone read exactly what runs; it says nothing about whether what runs is correct. The market contracts are not upgradeable proxies: what is verified is what will keep running.
Audited: the two bases, Synthetix StakingRewards and Compound V2, by their original auditors at the versions we forked.
Reviewed internally, not yet by a third party: our eighteen lines of staking changes; the risk engine, the lens, the benefits module and our oracles; the locker, the vault, the campaigns and the fee route; and how these components integrate.
Each property described above can be checked on chain today, independently of any audit, from the addresses in the contracts section.
This interface is built with AI assistance under human direction. What that changes is how fast it is written; what it does not change is where any number comes from. Every figure on the site is read from the contracts, and the contracts are what you should verify.