Understanding AMP

Every cryptocurrency has the same problem at a physical register. Bitcoin takes about 60 minutes to fully settle on-chain. Ethereum takes roughly 13 minutes to reach finality (the point where a transaction is irreversible) under Casper FFG, its consensus mechanism.^[12] Even faster chains need several confirmations before a payment can be called final. Merchants can't hold a customer at the checkout for that long, and they can't accept a payment that might still reverse.

AMP solves this by acting as collateral. When you pay with crypto through Flexa, AMP is locked in escrow for the duration of that transaction's settlement window. The merchant receives confirmation in under a second, guaranteed by the locked AMP, not by waiting for block confirmations. Once the underlying blockchain payment completes, the AMP unlocks and re-enters circulation. If the payment somehow fails, the collateral covers the merchant. Either way, the merchant gets paid. Think of stakers as surety bond providers: they lock AMP to guarantee merchant payments settle, and earn transaction fees for taking on that risk.

The result is sub-second checkout at the register, backed by a guarantee that doesn't require trusting the customer, the blockchain, or a payment processor. Just the collateral math. Because AMP is the collateral layer rather than the payment asset itself, Flexa can accept any digital asset the customer holds (BTC, ETH, SOL, USDC, and 99+ more), all using the same guarantee mechanism.^[1]

Flexa launched in 2018 with a mission to make cryptocurrency spendable everywhere.^[3] In 2019, they introduced Flexacoin (FXC)^[4] — a standard ERC-20 (Ethereum's basic token format) that served as collateral for Flexa's payment network. It worked, and the system ran in production. But the architecture had a hard limit: to use tokens as collateral, they had to be transferred to a smart contract. That's custodial. Your tokens were no longer yours while they were locked.

To build something fundamentally better, two capabilities were required that FXC couldn't support:

Neither feature could be retrofitted onto FXC after the fact. A new contract was the only path. Before deploying a single token, two independent security audits were commissioned. ConsenSys Diligence reviewed the contract in June 2020. Trail of Bits ran a second audit in July–August 2020. Both came back without critical findings.^[6,7] The migration was 1:1: burn FXC, receive AMP.^[5]

The deeper difference: FXC couldn't assign conditional access rights to tokens — it could only transfer them. AMP's partition system inverted this. A single balance can be segmented into independently-governed sub-accounts, each with its own transfer rules enforced by the AMP contract itself via ERC-1820 hooks.^[2] A collateral manager can lock a partition and veto any release attempt until conditions are met — without taking custody of the tokens. The result: the same AMP can be reused across transactions once each settles, rather than requiring a separate token pool per lock — a structural capital efficiency advantage no escrow-based design can match. As of 2026, no comparable implementation exists in production — though Flexa's own staking moved to the Anvil Vault deposit model in Capacity v3 (covered below); the partition system remains in the immutable AMP contract, but it is no longer the path Flexa uses for collateral.

Standard ERC-20 contracts are ledgers. One balance per address. To use those tokens as collateral, they have to move: out of your wallet, into a contract you don't fully control. AMP changes this with a partition system built into the token contract itself.

Every AMP wallet can contain multiple sub-accounts called partitions, each identified by a 32-byte prefix. Your wallet might show 50,000 AMP total, but inside, 20,000 might sit in a default (liquid) partition and 30,000 in a Flexa collateral partition. Both are yours. The collateral manager can restrict which partition you withdraw from, but it cannot take the tokens or transfer them elsewhere.^[2]

The locking mechanism works through transfer hooks registered via ERC-1820 (an on-chain registry that lets smart contracts subscribe to token events, similar to webhooks in web development).^[2,9] When AMP moves between partitions, it calls pre- and post-transfer hooks on any registered collateral manager. Those hooks can approve or block the movement until conditions are met, like a blockchain payment reaching finality.

Capacity v3 (Anvil) changes the custody model: instead of partition-based locking within your wallet, tokens are deposited into the Anvil Vault contract. Your AMP leaves your wallet and is held by the vault smart contract, with a 12–24 hour cooldown on withdrawals. The partition architecture remains part of the AMP contract, but the current staking interface uses the vault deposit model.

Most staking tokens generate yield by printing more tokens. AMP doesn't. The economics work differently and that distinction matters.

Merchants on the Flexa network pay roughly 1% per transaction,^[11] a fraction of the 3–4% charged by major card networks. Per the AMP whitepaper design, Flexa uses merchant fee revenue to buy AMP on the open market for distribution to stakers.^[2] Those purchased tokens are then distributed to stakers pro rata, based on the size and duration of their collateral stakes. No new tokens are minted. The yield comes from actual commerce.

As stakers lock up AMP to earn yield, the liquid circulating supply contracts. Less liquid supply with continuous buy pressure from the fee mechanism creates compounding price dynamics. More transaction volume means higher staking yields, which incentivizes more staking, which tightens supply further. The loop is self-reinforcing as long as Flexa's network grows.

It's not a governance token — AMP holders don't vote on protocol parameters. It's not inflationary staking — no new tokens are minted to pay stakers. It's not a wrapped asset — there's no underlying BTC or ETH behind it. It's collateral, a category that operates by different rules from most of what's out there.

100 billion tokens is an unusual number. It can feel like a red flag for people accustomed to assets with hard scarcity baked in. The context: AMP's supply came from FXC's original issuance and migrated 1:1. It wasn't designed for individual token scarcity. It was sized to provide enough granularity for fractional collateral across millions of simultaneous transactions.

At AMP prices around $0.0005 (mid-2026), collateralizing a single $50 transaction at 1.5x overcollateralization requires around 150,000 AMP. Scale that across thousands of concurrent transactions happening simultaneously and you need a large token count to make the math work at any meaningful payment volume.

The more relevant number is the accessible float. Approximately 20.5 billion AMP, around 21% of supply, was staked in Flexa Capacity as of mid-2026.^[14] That locked supply doesn't trade. As the network grows and staking yields improve, that percentage is expected to climb. Total supply is hard-capped at 100 billion — the only new AMP comes from the last sliver of unswapped Flexacoin converting 1:1. The liquid float is what moves.

AMP is the collateral engine. Flexa is the network it powers. The table below scores nine live payment rails across 11 criteria: settlement architecture, chargeback immunity, finality speed, payment guarantee, privacy, fees, refunds, fiat payout, POS reach, e-commerce coverage, and agentic compatibility.

Criterion 4, Payment Guarantee, is where Flexa separates from every competitor. It's the only rail with a staked collateral backstop: AMP is slashed to make the merchant whole if a payment fails pre-confirmation. No other scored rail replicates this. The other differentiator: Flexa accepts any digital asset the customer holds (BTC, ETH, SOL, USDC, and 99+ more) because the collateral math works the same regardless of which chain settles the underlying payment.

A lot of crypto whitepapers describe things that never happen. AMP is one of the cases where the core claims are largely in production. The quick version: sub-second checkout is real, the fee-to-staker mechanism is running, and the token contract has been live and immutable for over four years. Geographic scale is the honest gap.

Capacity v3 introduced ANVL as the protocol governance token. ANVL holders vote on pool parameters — which wallet apps qualify, how rewards are weighted, and how fee structures are set. If you encounter references to ANVL outside this site, that is the governance system for the staking pools described above.