What is Aptos' AIP-146 and Is It Good News for APT?

AIP-146 lets anyone who controls enough staked $APT push transactions with up to 100 times the network's normal gas limits, and on balance it reads as good news for the token: it gives the biggest players in DeFi a practical reason to stake APT beyond yield alone. This is not a distant roadmap item either. The proposal has been accepted, an onchain governance vote to enable it passed with near-unanimous support, and Aptos Labs (@AptosLabs) began promoting the upgrade this week.

What Is AIP-146?

Every Aptos transaction currently runs under the same execution and I/O gas limits. That works for swaps and transfers, but it caps the heavy stuff: DEX liquidations, risk engine updates, large state migrations, and complex financial logic that needs more compute in a single transaction. Teams that hit the ceiling either split the work into pieces or move parts of it offchain.

AIP-146, titled Staking-Based Transaction Limits and authored by George Mitenkov (@GeorgeMitenkov) of Aptos Labs, adds an opt-in fix. A transaction can carry a request for higher limits, expressed in basis points, from just above 1x to a hard cap of 100x. To qualify, the fee payer must prove control over substantial committed APT stake. Only execution and I/O limits can be raised in this first version, while storage fees stay unchanged.

The privilege comes with a built-in cost. Any staking-backed high-limit transaction must pay at least 10 times the minimum gas unit price, a deliberate brake on spam and frivolous use. High-limit transactions also get no special treatment in the mempool.

How Do the Staking Tiers Work?

Three routes count as staking proof: owning a stake pool, being the delegated voter of a stake pool, or holding stake in a delegation pool. Committed stake means active plus pending-inactive balances, checked against live onchain state.

The genesis configuration sets three tiers, with separate thresholds for execution and I/O:

• 2x multiplier: 1 million APT (execution) or 5 million APT (I/O)
• 4x multiplier: 5 million APT (execution) or 10 million APT (I/O)
• 8x multiplier: 10 million APT (execution) or 20 million APT (I/O)

I/O thresholds run higher because state reads and writes are more expensive for the network than raw compute. Requests round up: ask for 3x and you need the stake for the 4x tier. Governance can add, remove, or reshape tiers at any time without code changes, so the 100x ceiling is headroom for the future rather than something available on day one.

Will AIP-146 Lock Up More APT?

This is where the proposal gets interesting for APT holders. Access to higher limits is valuable for any protocol running serious onchain workloads, and the only way in is committed stake. That pulls DEXes, perps platforms, market makers, and institutions toward staking or arranging stake, on top of the usual draws of rewards and governance power.

Aptos has spent recent months reworking its tokenomics: a hard supply cap of 2.1 billion APT against a current supply of roughly 1.2 billion, transaction fee burns, and a base staking reward rate cut through governance to about 2.6%. With yield alone a weaker draw than it used to be, tying network capability to stake gives large holders a fresh incentive to lock APT rather than keep it liquid. Roughly 770 million APT is already staked, around 64% of supply, so the mechanism builds on an already high baseline.

The rollout has moved quickly. Authored in April 2026 and accepted in the AIP repository in early July, the proposal was enabled through onchain proposal 200, Enable transaction limits. That vote drew 369.9 million APT in favor and just 1,449 APT against, on a 30.7% turnout, and has since been executed on mainnet. Practical adoption will still lag: SDK and CLI support is tracked separately, and wallets need to add the new transaction format before high-limit transactions become routine.

Is It Good News for APT?

Mostly, yes. The upgrade removes a technical ceiling for exactly the kind of applications @Aptos has been courting with its onchain markets push, and it does so in a way that converts network capability into staking demand. Large stakers carrying real economic weight also have skin in the game, which is the logic of tying privileges to stake.

The entry ticket starts at 1 million APT, so this is a feature for protocols, funds, and whales, not retail. It concentrates a network privilege among the largest holders by design. The mechanism ships as an experiment, with tiers and access subject to change by governance, so builders are told not to depend on guaranteed access. Still, the direction is clear: Aptos just handed its heaviest users a real reason to lock APT, and the staking rate will show whether they take it. 

Sources:

• AIP-146 Full proposal text with tier parameters, validation rules, and security considerations.
• Aptos Labs Announcement post on X, July 21, 2026.
• GovScan Onchain voting data for proposal 200, which enabled the feature on mainnet.
• Aptos Network Official APT supply, emissions, and burn dashboard.
• AIP-146 Discussion Community discussion thread on GitHub.