Scan DEX pairs for MEV & sandwich risk

A sandwich attack is a three-transaction MEV pattern. A bot watching the mempool spots your pending swap, places a buy right before it to push the price against you, lets your trade execute at the worse price, then sells right after to pocket the difference. It only works on AMM-based DEXs, where trade size and slippage tolerance let the bot calculate its profit ceiling from your transaction.

It returns a query, not live results. The builder assembles a ready-to-run config (mode, chain, token or pool addresses, max_results) and shows a fixed example of the output shape so you know what fields to expect. The underlying source is anti-bot and not CORS-open, so live scanning happens when you run the query on the backing Apify actor, not in your browser.

Per pool you get sandwich_rate_24h (share of transactions sandwiched), estimated_sandwich_attacks_24h (a 24-hour count), mev_risk_level (low/medium/high or null), mev_to_total_fees_ratio_24h (how much of fees flow to MEV builders), and fee_regime_24h (mev-dominated, gas-dominated, or pool-fee-dominated). A fees_24h object splits total cost into base fees, priority fees, builder tips, pool fees (labelled 'Pool fees (USD)', i.e. LP fees), and L1 data fees. Coverage is not uniform: builder tips are genuinely zero on L2s like Base/Arbitrum/Optimism/Polygon (so mev_to_total_fees_ratio_24h is then null), L1 data fees are null on L1 chains, and Solana coverage is per-pool — some pools return only the sandwich rate. Every row includes a data_availability map showing which fields populated for that pool.

The backing actor queries Codex.io's GraphQL index, which covers 133 blockchain networks — Ethereum, Solana, Base, BSC, Arbitrum, Optimism, Polygon, Avalanche, Sui, Aptos, and many more. Bare 0x addresses resolve via the default_evm_chain dropdown (defaults to Ethereum), Solana base58 addresses auto-detect, and an explicit chain:address prefix overrides the default per row.

On Solana, bots reportedly drained a large sum from users (figures circulating in 2025 crypto-research write-ups range up to roughly $500 million over a multi-month period ending in 2025), before coordinated mitigation — validator action and Jito closing its public mempool — cut profitability substantially. On Ethereum, community MEV-data dashboards have tracked tens of thousands of sandwich attacks per month through 2025, often at a small average profit per attack. Treat these as third-party estimates, not on-chain ground truth.

Use scan-token to discover and scan a token's top-volume pools when you only know the token. Use scan-pair when you already have specific pool/pair addresses and want direct analysis. Use discover-mev to rank the most MEV-toxic pools across selected chains by risk level, filtered by a minimum-liquidity threshold — useful for finding which pools to avoid before you trade.

Common defenses are setting a tight slippage tolerance (around 0.1% on stable pairs, 0.5-1% on liquid pairs), routing through a private RPC such as Flashbots Protect or MEV Blocker so your transaction skips the public mempool, and using MEV-protected aggregators like CoW Swap or 1inch Fusion. This scanner helps the planning side: spotting high-sandwich-rate pools and, when available, lower-MEV trading windows before you swap.

The in-browser query builder is free. Running the actor on Apify is free to start using your Apify platform free credits — there is no separate actor-specific free quota — and then pay-as-you-go. The actor uses pay-per-event pricing: a per-run start fee, a per-pair-scanned charge for each pool analyzed with full MEV intel and 24h hourly bars, and an optional small charge per AI verdict if you enable the trader summary. You only pay for the pairs you actually scan.

Sometimes, when the data supports it. Each scanned pool includes safer_trading_windows_utc — the hours over the last 24 hours where the pool's MEV risk was 'low', derived from bars_hourly time-series data (OHLCV plus MEV and fee columns). Note this field is frequently empty: on persistently MEV-dominated pools — often the very high-risk pools you'd most want to time — there may be no low-risk hour, and the array comes back empty. When it does populate, combine it with sandwich_rate_24h and the fee regime label to time a trade into a lower-toxicity window.