Aptos activated an opt-in encrypted mempool on mainnet on October 12, 2026, becoming the first major Layer 1 blockchain to embed transaction-payload encryption directly into its validator set. The feature, specified in Aptos Improvement Proposal 144 (AIP-144) and authored by Aptos Labs cryptograp...
Aptos activated an opt-in encrypted mempool on mainnet on October 12, 2026, becoming the first major Layer 1 blockchain to embed transaction-payload encryption directly into its validator set. The feature, specified in Aptos Improvement Proposal 144 (AIP-144) and authored by Aptos Labs cryptographer Rex Fernando, uses batched threshold encryption to hide transaction payloads — including called functions, input parameters, asset types, and transfer amounts — until a block is finalized. The minimum gas-unit price for encrypted transactions doubles to 200 octas, with an additional 375 gas units charged for payload decryption.
The launch coincides with the completion of Aptos's four-year genesis token unlock schedule. As of October 12, all 190 million APT allocated to core contributors and approximately 134.78 million APT to early investors have fully vested. Monthly token emissions dropped approximately 60%, from 11.31 million APT to 4.54 million APT. APT traded near $0.90 at the time of the announcement.
The encrypted mempool addresses a quantifiable economic problem. Aggregate MEV extracted across all major chains exceeds $7.2 billion since 2020, according to Flashbots and Jito data. Ethereum alone has seen cumulative MEV surpass $4 billion through 2026, with daily extraction running $10–20 million on normal days and spiking to $40–50 million during periods of volatility. Sandwich attacks account for roughly 30% of all Ethereum MEV — approximately $2.2 billion lifetime. On Solana, aggregate searcher profits captured through Jito bundles passed $480 million cumulative as of Q2 2026. The cost to individual traders on default settings runs 0.05% to 0.5% per trade on volatile pairs.
The system operates through a threshold encryption scheme integrated into Aptos's existing AptosBFT consensus mechanism. The flow proceeds as follows:
Client-side encryption. A compatible wallet encrypts the transaction payload on the sender's device before submission. Sender addresses and gas fees remain visible; the called function, input parameters, asset types, and transfer amounts are encrypted.
Blind ordering. Validators receive the encrypted transaction and include it in block proposals. They order transactions without access to payload contents. This eliminates the information asymmetry that MEV bots exploit — validators cannot see what a transaction does until after sequencing is final.
Threshold decryption. The decryption key is split across validators by stake weight. A single validator, or any group representing less than two-thirds of the network's staked voting power, cannot decrypt pending transactions independently. Only after a block is finalized do validators collectively decrypt payloads for execution.
Execution. Decrypted transactions execute immediately after block finalization. If a transaction is not finalized, it will not be decrypted under the honest-majority validator assumption.
The feature is opt-in. Users and applications can choose whether to encrypt each transaction. Standard unencrypted transactions continue to function as before.
Encryption is not free. Aptos doubled the minimum gas-unit price for encrypted transactions to 200 octas (compared to 100 octas for standard transactions) and added 375 gas units for payload decryption overhead.
The practical cost impact is significant for small transactions. According to Aptos documentation, a standard transfer using 150 gas units costs 15,000 octas at the base 100-octas-per-unit rate. The same transfer encrypted costs approximately 105,000 octas (0.00105 APT) — roughly seven times the unencrypted bill. At current APT prices near $0.90, the encrypted transfer costs approximately $0.0009, which remains negligible in absolute terms. The cost differential matters more as a percentage of small-value on-chain operations.
Performance overhead is minimal. Testing across a 50-node environment measured approximately 20 milliseconds of added latency per block, which falls within Aptos's sub-second finalization window under AptosBFT consensus.
ZKSecurity conducted an independent security audit of the encrypted mempool prior to mainnet deployment. According to the published audit report, the review was divided into distinct phases covering:
Phase 1: Integration of batched Identity-Based Encryption (IBE) threshold encryption, integration of Chunky Distributed Key Generation (DKG), decryption secret key share derivation and aggregation, and the full transaction encryption-proposal-validation-decryption-execution lifecycle.
Cryptographic evaluations: Whether the batched IBE threshold encryption scheme provides CCA (Chosen-Ciphertext Attack) security, whether the Chunky PVSS (Publicly Verifiable Secret Sharing) scheme correctly distributes secret shares among validators, and whether the DeKART zero-knowledge range proof is sound.
The cryptographic design relies on a batched IBE threshold encryption scheme, meaning transactions are encrypted under an identity derived from the upcoming epoch, and decryption requires a threshold of validator key shares. This avoids the need for a separate encryption committee or trusted third party — the existing Aptos validator set handles both ordering and decryption.
At launch, the following integrations are confirmed or announced:
The developer integration path is notable: applications do not need to rewrite contracts to support encrypted transactions. The encryption layer operates at the transaction submission level, below the smart contract execution layer.
The encrypted mempool is a response to a well-documented economic extraction problem across blockchain networks:
| Metric | Value | Source | |--------|-------|--------| | Cumulative MEV, all chains since 2020 | >$7.2 billion | Flashbots, Jito | | Cumulative MEV, Ethereum alone | >$4 billion | Flashbots | | Ethereum daily MEV (normal) | $10–20 million | Flashbots | | Ethereum daily MEV (volatility) | $40–50 million | Flashbots | | Sandwich attacks, % of Ethereum MEV | ~30% | Flashbots | | Sandwich attacks, Ethereum lifetime | ~$2.2 billion | Flashbots | | Solana cumulative MEV (Jito bundles, through Q2 2026) | >$480 million | Jito | | Solana monthly MEV run-rate | $30–50 million | Jito | | Cost to individual traders per trade | 0.05%–0.5% | Academic estimates |
Monthly Ethereum MEV runs approximately $180 million; Solana generates approximately $45 million per month. On Solana, more than 78% of mainnet stake runs the Jito-Solana client, and Jito bundles account for the majority of intentional MEV extraction on the network.
The economic argument for encrypted mempools rests on redirecting this value from MEV searchers back to users. Whether encryption eliminates MEV or merely shifts extraction methods remains an open question. Academic research, including a paper titled "No Place to Hide: An Analysis on Protected Order Flow Sandwich Attacks" published in June 2026, indicates that private order flow is not immune to exploitation through more sophisticated statistical methods.
Aptos is not the only network addressing MEV through transaction privacy, but it is the first to embed encryption at the protocol layer in a major L1.
Ethereum does not have a protocol-native encrypted mempool. MEV mitigation relies on third-party infrastructure:
Ethereum's Glamsterdam upgrade (tested on Sepolia in October 2026 with 200M gas limit and ePBS) addresses block production structure but does not include encrypted mempools. Proposals for protocol-level encryption remain in the research stage.
Solana relies on Jito's block engine, which functions as a private ordering layer. Jito does not encrypt transactions but batches them to reduce sandwich attack viability. No protocol-native encryption exists or has been proposed for Solana's validator client.
The key architectural distinction: Aptos uses its existing validator set for both block ordering and threshold decryption. Ethereum's approaches require separate infrastructure layers (Flashbots relayers, Shutter keyper committees). This reduces trust assumptions on Aptos but concentrates decryption authority in the same validator set that orders transactions.
The encrypted mempool launch coincided with a structural change in APT token supply dynamics. On October 12, 2026, Aptos completed its four-year token unlock schedule for core contributors and early investors:
Post-vesting, monthly APT emissions dropped approximately 60%, from 11.31 million tokens to 4.54 million tokens per month. No further genesis insider token unlocks are scheduled. Ongoing distribution shifts to community and foundation pools.
As of mid-October 2026, Aptos network metrics provide additional context for evaluating the encrypted mempool's potential impact:
The relatively low DeFi TVL limits the immediate addressable MEV on Aptos compared to Ethereum or Solana. However, the feature positions the network for future growth if DeFi activity increases, and the Decibel integration suggests derivatives trading — where front-running costs are highest — is the initial target market.
Aptos's encrypted mempool represents a concrete engineering response to MEV extraction, with measurable costs (7x gas premium, 20ms latency) and clear cryptographic underpinnings (threshold IBE, ZKSecurity audit). The feature distinguishes itself architecturally from Ethereum's third-party solutions by operating within the protocol's existing validator infrastructure.
Whether this matters depends on adoption. The opt-in model means the feature's effectiveness scales with usage — if most transactions remain unencrypted, MEV bots can still extract value from the visible transaction pool. Aptos's current DeFi TVL of $50.5 million means the addressable MEV is small relative to networks where the problem is most acute. The Decibel integration, targeting perpetual futures where front-running costs are highest, suggests Aptos is prioritizing use cases where encrypted ordering delivers the clearest economic benefit.
The broader question for the industry is whether encrypted mempools will become standard infrastructure across L1s, or whether third-party solutions (Flashbots, MEV Blocker, Shutter) will remain sufficient. Aptos has the advantage of shipping first at the protocol layer. Whether other networks follow — and whether academic research on statistical exploitation of private order flow proves these systems ultimately insufficient — remains to be determined.