Sui Move MEV 2026: Parallel-Execution Searcher Playbook

Answer first — MEV on Sui in 2026 is structurally different from EVM chains because of two architectural choices: object-centric ownership (most transactions touch user-owned objects and don't conflict, so they execute in parallel) and Mysticeti DAG consensus (no single linear mempool, no single block-builder auction). For searchers this means: classic sandwich attacks are mostly impossible against single-owner-object transactions; the live opportunities are shared-object arbitrage (DeepBook, Cetus, Turbos, Aftermath pools), liquidation auctions on Sui lending, and package-publish back-running when new contracts deploy.

Why Sui Breaks The EVM MEV Mental Model

EVM MEV exists because:

1. Transactions queue in a public mempool
2. Block proposers (or builders) order transactions however they want
3. Sequential execution makes ordering profitable to manipulate

Sui changes all three:

1. Owned-object transactions skip consensus and execute directly — no mempool ordering opportunity
2. Shared-object transactions go through consensus but are executed in parallel where possible
3. Mysticeti is a DAG, not a single chain — there's no single "next block" to bribe a builder for

The result: roughly 60–80% of all Sui transactions (transfers, NFT operations, owned-object updates) are immune to traditional MEV because they never touch a shared resource that could be sandwiched.

The remaining 20–40% (DEX swaps, lending operations, perp positions, shared object writes) are where MEV lives. The opportunity surface is smaller than EVM but the competition is also thinner.

Opportunity 1: Shared-Object DEX Arbitrage

DEX pools on Sui are shared objects. When two users swap on the same pool in the same checkpoint, they're serialized through consensus. This serialization creates ordering opportunities similar to EVM blocks.

The major venues in 2026:

Cross-DEX arb works the same way it does on EVM: monitor prices, detect spreads exceeding gas+fees, submit atomic bundles. The chain's parallel execution means your arb transaction can run in the same checkpoint as the swap you're arbing, not the next block.

Per-trade profit on the most active pairs (SUI/USDC, ETH/USDC, BTC/USDC, AFSUI/SUI): $5–$300. The competitive landscape in early 2026 has roughly 8–15 active searchers — competitive but not saturated like Ethereum.

Opportunity 2: DeepBook Orderbook MEV

DeepBook is Sui's native CLOB. Unlike AMM-based MEV, orderbook MEV maps better to traditional HFT:

• Maker-taker rebates reward fast quoting
• Order racing on resting orders that suddenly become profitable
• Stale-quote arb between DeepBook prices and AMM prices

DeepBook's per-pair fee structure and Mysticeti's low-latency consensus make some patterns viable that wouldn't work on EVM CLOBs. Specifically: sub-second quote updates are real, not theoretical. A bot that updates quotes 4-8 times per second can earn maker rebates that a slower bot cannot.

This is closer to traditional electronic market-making than to EVM searcher work. Different skill set — but the same person can do both with different tooling.

Opportunity 3: Liquidation MEV

Sui lending markets — Suilend, NAVI, Scallop — run liquidation auctions when positions go under-collateralized. Liquidator bonus ranges 5–10% of seized collateral.

Differences vs EVM liquidation MEV:

• Parallel execution: Multiple liquidators can attempt the same liquidation in the same checkpoint without one transaction blocking the others — the protocol's first-come-first-serve logic determines the winner, not the mempool order.
• Oracle latency: Sui's price oracles (Pyth, Switchboard) update on-chain at sub-second cadence. Liquidation triggers can fire faster than on slower chains.
• Inventory in SUI: You need SUI for gas and a reserve of borrowable assets to repay the loan. Inventory management is different from holding ETH/USDC on mainnet.

See Liquidation MEV Strategy 2026 for the cross-chain framework. The Sui-specific tweak is mostly: less ordering manipulation, more pure execution speed.

Opportunity 4: Package-Publish Back-Running

When a new Move package (contract) is published on Sui, the first transactions interacting with it often establish prices, liquidity, or initial state. Searchers monitoring the Publish transaction type can back-run new pool deployments to capture initial price discovery.

This is analogous to launch sniping on Solana but with a different signature pattern. On Sui, the publish includes the package bytecode — a searcher can parse it for the new contract's interfaces and pre-build interaction transactions before any human can.

Realistic returns: 10–50% on initial liquidity additions where the deploy price differs significantly from any reference oracle. Not a high-frequency strategy — maybe 1–3 opportunities per day during active launch periods.

See Solana Pump.fun Sniper Bot Guide for the launch-sniping mental model; the Sui variant is structurally similar.

What MEV Does Not Work On Sui

Three EVM patterns that do not translate:

Sandwich attacks on user-owned transfers

A SUI transfer between two wallets is an owned-object operation. It bypasses consensus ordering entirely. There's no sandwich opportunity because the transfer doesn't interact with a shared object that could be reordered around.

Generic mempool front-running

There's no single mempool. Transactions submitted via different validators may not see each other until consensus. The "watch mempool, race to include before the target" pattern doesn't apply.

Block-builder bribery

There are no PBS-style builders on Sui. Validators commit blocks through Mysticeti's leader rotation; there's no auction market for inclusion ordering analogous to Flashbots or MEV-Boost.

This is why searchers migrating from Ethereum often fail on Sui for the first month: their entire mental model is wrong. Sui MEV is more like CEX market-making with parallel execution than mainnet searcher work.

Tooling Gaps In 2026

Honest assessment of what's missing:

• No mature private orderflow on Sui — most arb is fought in public consensus. This isn't a problem yet because competition is thin, but it will be once volume grows.
• Limited indexer/RPC options — Sui's RPC ecosystem is smaller than Ethereum's. Most serious searchers run their own full node.
• Move tooling is younger than Solidity tooling. Simulation, decompilation, and debugging are workable but not at parity with Foundry/Hardhat.
• MEV research is sparse — much less academic and Flashbots-style published research on Sui MEV than on EVM. You're partly writing the playbook yourself.

This is the trade-off of working on a less-crowded chain: more raw alpha, fewer ready-made tools. See Best WSS Endpoints by Chain 2026 for the broader infrastructure context.

Realistic Returns

Indicative early-2026 monthly returns for a solo searcher with $20k working capital on Sui:

• Cross-DEX AMM arb: 3–8% monthly
• DeepBook market-making + stale-quote arb: 5–15% monthly, requires more sophisticated tooling
• Liquidation MEV: 2–6% monthly, event-driven
• Launch sniping new pools: highly variable, can be 20%+ in a month with multiple launches

These ranges are illustrative, not promises. The returns are noisier than mature EVM chains because the per-event sizes are smaller but more frequent. See the FRB risk disclosure for the full risk model.

What FRB Agent Supports

FRB Agent does not currently support Sui in its public release. The EVM atomic-arbitrage engine doesn't translate to Move's object semantics, and Sui-specific support would require a separate execution engine. FRB Agent's primary Solana support (via the dedicated Solana app) handles the parallel-execution paradigm but in SVM, not Move.

Searchers running on Sui today typically use:

• Custom Move scripts compiled to publish bundles
• Self-hosted full nodes for low-latency mempool/checkpoint observation
• Their own indexers built on top of sui-indexer or community alternatives

If Sui support becomes a priority, FRB Agent could add a Move-execution module — but as of early 2026, it does not.

Further Reading

• Liquidation MEV Strategy 2026
• Solana Pump.fun Sniper Bot Guide
• Monad MEV: Parallel EVM 2026
• Best WSS Endpoints by Chain 2026
• Best Chain for MEV 2026