DeepChain深链
Layer 1 explainer 06 / 08  |  aptos  |  apt  |  4 min read
APT

Aptos: Diem's other heir

Like Sui, it comes from Meta's Diem team and uses the Move language, but it takes a different route: transactions are executed optimistically in parallel and re-run if they conflict. Blocks are extremely frequent, with fewer than a hundred validators.

Layer 1 · AptosData as of 16 Oct 2026, 01:45 UK timeWebsite aptosfoundation.org ↗
Block time
36 ms
Validators
83

In one sentence

After Meta’s stablecoin project Diem was wound up in 2022, its core team split in two: one group built Sui, the other built Aptos. Both chains use Move, but they made different design choices:

  • Sui changed the data model (everything is an object) so that non-conflicting transactions simply skip the queue.
  • Aptos kept an account model closer to traditional blockchains and focused on execution: assume transactions do not conflict, run them all in parallel, and redo them if a conflict turns up.

Where it came from

  • Late 2021: former Diem leads Mo Shaikh and Avery Ching founded Aptos Labs and went on to raise several large rounds.
  • October 2022: Aptos mainnet launched and APT began trading at the same time. Early on there was little on-chain activity and large team and investor allocations, which drew plenty of criticism.
  • The following years: the focus shifted to stablecoins, real-world asset tokenisation (RWA) and payments, and several institutions deployed tokenised products on Aptos.

Parallel execution: Block-STM

A block holds hundreds or thousands of transactions. The safest approach is to execute them one by one in order, but that is slow. Block-STM works like this:

  1. The order of transactions in the block is fixed first.
  2. All transactions run at the same time on multiple CPU cores, each recording what data it read and what it changed.
  3. If a transaction turns out to have read data that an earlier transaction in the order changed, it is discarded and re-run.
  4. This continues until every result matches what sequential execution would have produced.

When most transactions are unrelated, this “do it first, check later” approach uses every core; when there are many conflicts (say everyone is hitting the same pool), it degrades into repeated retries.

Aptos blocks are small and frequent: measured, a block arrives on average every 36 ms, and there are 83 active validators.

how it worksHow a transaction gets on-chain
  1. 01Signed and sent

    The wallet signs and sends it to a full node. Fees are paid in APT.

    Instant
  2. 02Into the mempool

    The transaction is forwarded to validators and waits to be ordered into a block.

  3. 03Consensus ordering

    Validators use AptosBFT to agree on the block's contents and order.

    About 36 ms per block
  4. 04Block-STM parallel execution

    Transactions run in parallel, conflicting ones are re-run, and the result matches sequential execution.

  5. 05Finalised

    Irreversible once validators with over two-thirds of stake vote to confirm.

    Usually under a second

APT and fees

Fees on Aptos are paid in APT; APT is also staked with validators and used in governance. Under the current rules fees are burned, and validators earn mainly from staking rewards (newly issued APT).

About 883.01M APT are in circulation, a market cap of $739.3M. Team and investor allocations unlock monthly on a schedule, creating persistent selling pressure.

What is on-chain today

live dataData as of 16 Oct 2026, 01:45 UK time
Token market cap$739.3MMarket cap rank #93
TVL$44MAssets in DeFi protocols
Stablecoins$885.8MUSDT, USDC and others combined
24h DEX volume$5.4M30 days $186M
24h fees$5,45630 days $137K
Block time (measured)36 msaverage of the last 10K blocks
Validators / block producers83
Circulating supply883.01Mcap 2.1B
Stablecoin mix total $885.8M
  • USDT 47% $416.6M
  • USDC 33% $292.5M
  • Other 20% $176.6M
Top 6 protocols by TVL TVL on this chain
  1. DecibelDerivatives$29.6M
  2. KAIORWA$6.6M
  3. Echelon MarketLending$6.6M
  4. Amnis FinanceLiquid Staking$4M
  5. Thala LSDLiquid Staking$1.9M
  6. TruStakeLiquid Staking$1.7M

Sources: CoinGecko (market cap, supply), DefiLlama (TVL, stablecoins, DEX volume, fees, protocols), measurements from public nodes (block time, validators).

Things worth noticing:

  • Far more stablecoins than DeFi: $885.8M of stablecoins against only $44M locked in DeFi. Money mostly sits on the chain for transfers and payments rather than in DeFi.
  • Locked value is concentrated in a few protocols: one or two each in derivatives, RWA, lending and liquid staking.

Strengths and trade-offs

Strengths
  • Frequent blocks and fast confirmation; parallel execution makes full use of multi-core hardware.
  • Move inherently prevents assets from being copied or lost out of thin air, making one class of common bugs harder to write.
  • Strong presence in stablecoins and institutional assets (RWA).
Trade-offs
  • Fewer than a hundred validators; decentralisation is limited.
  • Not EVM-compatible, so developers and users have to adapt.
  • Small DeFi and modest ecosystem activity; large team and investor allocations keep unlocking.
Investigating a project on Aptos? Check these first
  1. 01

    Which token standard it uses (the older Coin or the newer Fungible Asset), and who holds the mint, freeze and burn rights (MintRef / FreezeRef / BurnRef).

  2. 02

    The package's upgrade policy: only packages set to immutable cannot change; all others can be upgraded by the publisher.

  3. 03

    Among the top ten holders, exclude exchanges, pools and staking contracts first, then check whether the rest come from the same source.

  4. 04

    Aptos and Sui addresses look very similar, so make sure you picked the right network before sending.

Data and sources

  • Live figures: CoinGecko, DefiLlama and the official Aptos node API (measured block interval, active validators).
  • Mechanisms are based on the official aptos.dev documentation and the Block-STM paper.
  • This guide is educational and is not investment advice.