MEV Risk Scanner — DEX Sandwich + Fee Intel

Per-pair MEV exposure, sandwich-attack rate, and fee-regime classification across 133 chains. Powered by Codex.io.

The first Apify actor that surfaces MEV intelligence per DEX pair. While other crypto scrapers stop at price + volume, this one measures what your trade actually costs — gas vs builder tips vs LP fees, MEV-to-total-fees ratio, sandwich-attack frequency, and "safer trading windows" derived from hourly time series. Built on Codex.io, with optional AI verdicts via Gemini for non-technical traders.

Why MEV intelligence matters

In 2025, sandwich attacks extracted ~$500M from Solana traders alone, and individual high-profile losses on Ethereum reached six figures (the March 2025 stablecoin-swap loss of $215K on Uniswap V3 is one example). Yet no Apify actor exposes the underlying data — mevRiskLevel, sandwich rate per transaction, or fee decomposition — that would let traders avoid those pools or time their trades.

This actor closes that gap. For a given pair, you see:

• Whether it's MEV-dominated, gas-dominated, or pool-fee-dominated
• The % of fees that go to MEV builders (Ethereum) or Jito tips (Solana)
• The sandwich-attack rate per transaction and an estimated 24h sandwich count
• A complete fee breakdown: gas, priority fees, builder tips (MEV), pool/LP fees, L1 data fees (L2 rollups only)
• 24 hourly bars of OHLCV + the same MEV/fee fields per bar — pinpoints which hours are safest to trade

Why use this actor

• Unique data on Apify — no other actor exposes MEV/fee/sandwich fields. Codex's API is the only consumer-accessible source for these and this is the only Apify wrapper.
• 133 chains — Ethereum, Solana, Base, BSC, Arbitrum, Optimism, Polygon, Avalanche, Sui, Aptos, HyperEVM, Monad, and 120+ more. Cross-chain comparisons in a single run.
• Three modes in one actor — scan specific pairs you care about, scan a token's top pools, or discover the most MEV-toxic pools chain-wide.
• Per-row data quality flags — every output row includes a data_availability map showing which fields populated for that pool. No false zeros.
• Optional AI verdict (Gemini Flash) — turns the raw numbers into one short trader-language sentence per pair: "Medium MEV, 24% of fees are MEV, avoid this pool unless using a private mempool."
• Bulk CSV / JSON export — works with everything Apify already supports: scheduled runs, webhooks, integrations.

Modes

scan-pair — direct DEX pool scan

You provide a list of pair addresses; we return MEV intel + 24h hourly bars for each. Use when you already know which pool(s) to check (e.g. before a large trade).

scan-token — top pools for a token

You provide a token address; we discover its top-volume DEX pools (filtered by liquidity), and scan each. Use to compare MEV exposure across a token's pools — same token / different pools often have very different MEV profiles.

discover-mev — find the most MEV-toxic pools

We scan top-volume pools across the chains you specify and rank them by MEV risk (high → medium → low) then MEV-to-fees ratio. Use as a daily MEV-leaderboard or to flag emerging high-toxicity pools.

How to use

1. Click Try for free and open the input form.
2. Pick a mode:
   • Scan a token's top pools (recommended starting mode) — paste any ERC-20 or SPL token address.
   • Scan specific DEX pools — paste one or more pool addresses.
   • Discover the most MEV-toxic pools — pick chains and a min-liquidity floor.
3. Address format is forgiving — you don't need to type a chain prefix:
   • Bare 0x... addresses use the Default chain for bare 0x addresses dropdown (defaults to Ethereum). Switch it to base, arbitrum, optimism, etc. for a non-Ethereum EVM run.
   • Bare Solana base58 addresses (like DezXAZ8z7PnrnRJjz3wXBoRgixCa6xjnB7YaB1pPB263) auto-detect as Solana — no setting needed.
   • For mixed-chain runs, override per-row with chain:address (e.g. base:0x4200... next to ethereum:0xc02aaa...).
4. Optionally enable Add AI trader verdict for a 1-line trader summary per pool (charges an extra ai-verdict event).
5. Click Start. Most runs complete in 5–30 seconds.
6. Open the Dataset tab for results. Two views are configured: an Overview table with the MEV summary, and a Fee breakdown table with the full fee decomposition.

Quick start recipes

Copy any of these directly into the input form (switch to JSON tab) for an instant working run.

🟢 First-time test — minimum cost, instant feedback

{ "mode": "scan-token", "tokens": ["0xc02aaa39b223fe8d0a0e5c4f27ead9083c756cc2"], "max_pairs_per_token": 1, "max_results": 1 }

Scans the top WETH pool on Ethereum. ~1 pair-scanned event ≈ $0.05. Returns full MEV intel + 25 hourly bars.

🟡 Memecoin MEV check (Solana)

{ "mode": "scan-token", "tokens": ["DezXAZ8z7PnrnRJjz3wXBoRgixCa6xjnB7YaB1pPB263", "EKpQGSJtjMFqKZ9KQanSqYXRcF8fBopzLHYxdM65zcjm"], "max_pairs_per_token": 2, "max_results": 4, "enable_ai_verdict": true }

BONK + WIF top pools on Solana with AI verdicts. ~4 pair-scanned + 4 ai-verdict ≈ $0.28.

🔵 Base ecosystem scan (use the default chain dropdown)

{ "mode": "scan-token", "default_evm_chain": "base", "tokens": ["0x4200000000000000000000000000000000000006", "0x833589fcd6edb6e08f4c7c32d4f71b54bda02913"], "max_pairs_per_token": 2, "max_results": 4 }

WETH + USDC top pools on Base. Bare 0x addresses → resolved to Base via the dropdown.

🔴 Daily MEV-toxic leaderboard

{ "mode": "discover-mev", "discover_chains": ["ethereum", "solana", "base", "arbitrum"], "discover_min_liquidity_usd": 1000000, "max_results": 10, "enable_ai_verdict": true }

Top 10 most MEV-toxic pools across 4 chains, ranked high → low. ~10 pair-scanned + 10 ai-verdict ≈ $0.72. Schedule daily for a leaderboard.

🟣 Pre-trade safety check on a known pool

{ "mode": "scan-pair", "pairs": ["0xe0554a476a092703abdb3ef35c80e0d76d32939f"], "max_results": 1, "bar_resolution": "5", "enable_ai_verdict": true }

USDC/WETH Uniswap V3 0.05% pool, 5-minute bars (~289 bars), AI verdict. Run this before placing a large trade.

Address format rules

You paste	Actor uses
0xc02aaa39... (40-hex EVM)	The chain you picked in Default chain for bare 0x addresses (defaults to Ethereum)
DezXAZ8z7Pnrn... (Solana base58)	Auto-detected as Solana, regardless of the dropdown
chain:address (explicit)	The chain you typed — overrides the dropdown for that row

Mix any of these in the same input — the actor logs which chain it picked for each address so you can verify before reading the output.

Input reference

See the Input tab for the full form. Quick-start recipes are above; below is the full schema.

mode (required)

Which workflow to run: scan-token, scan-pair, or discover-mev. See the Modes section.

default_evm_chain

Default chain used when you paste a bare 0x... EVM address (no chain: prefix). Defaults to ethereum. Switch to base, arbitrum, optimism, etc. for non-Ethereum scans without needing per-row prefixes. Solana base58 addresses always auto-detect regardless of this setting.

Available values: ethereum, base, arbitrum, optimism, polygon, bsc, avalanche, linea, scroll, mantle, blast, zksync, metis, celo, fantom, cronos, moonbeam, moonriver, ronin, sei, monad, hyper-evm.

tokens (used by scan-token mode)

List of token addresses, one per row. Bare addresses use default_evm_chain for EVM, auto-detect for Solana. chain:address overrides per row.

max_pairs_per_token

How many of each token's most-active pools to scan. Default 3, max 20.

pairs (used by scan-pair mode)

List of pool addresses, one per row. Same address rules as tokens.

discover_chains (used by discover-mev mode)

Canonical chain names — one per row. The actor scans each chain's top-volume pools (filtered by liquidity below) and ranks them by MEV risk. 133 chains supported; common picks: ethereum, solana, base, bsc, arbitrum, optimism, polygon, avalanche.

discover_min_liquidity_usd

Minimum pool liquidity for discover-mev. Default $1M. Lower for memecoin scans, higher for blue-chip-only.

max_results

Total pools returned per run, hard cap 20. Each pool fires one pair-scanned event ($0.05).

bar_resolution

5 / 15 / 60 / 240 / 1D — granularity of the 24h time-series bars. 60 (1h) is recommended and returns ~25 bars/24h. Tighter resolutions (5, 15) surface intraday MEV spikes; 1D collapses to a 1–2 bar overview.

enable_ai_verdict

When true, adds a 1-line trader-language verdict per pool via Gemini Flash (extra ai-verdict event @ $0.02 per pool).

Chain names accepted

Beyond the EVM chains in the dropdown above: solana, sui, aptos, plus any of the 133 supported networks via raw Codex slug (e.g. tron, kaia, tac). Network IDs are looked up dynamically at startup, so newly-added Codex chains work without an actor update. The full list prints in your run log.

Output

Each dataset row is a NormalizedPair:

{
    "pair_address": "0xe0554a476a092703abdb3ef35c80e0d76d32939f",
    "chain": "ethereum",
    "network_id": 1,
    "exchange": "UniswapV3",
    "token0_symbol": "USDC",
    "token0_address": "0xa0b86991...",
    "token1_symbol": "WETH",
    "token1_address": "0xc02aaa39...",

    // 24h aggregates
    "volume_24h_usd": 76851555.5,
    "liquidity_usd": 2761557.8,
    "transactions_24h": 15589,

    // MEV intelligence (24h dominant)
    "mev_risk_level": "medium",                // low | medium | high | null
    "mev_to_total_fees_ratio_24h": 0.248,      // 24.8% of fees are MEV builder tips
    "fee_regime_24h": "mev-dominated",         // mev-dominated | gas-dominated | pool-fee-dominated
    "sandwich_rate_24h": 0.00179,              // 0.18% of transactions sandwiched
    "estimated_sandwich_attacks_24h": 28,

    "fees_24h": {
        "total_usd": 28820.5,
        "base_fees_usd": 4844.5,        // gas
        "priority_fees_usd": 8798.3,
        "builder_tips_usd": 7503.5,     // MEV bribes
        "pool_fees_usd": 7674.1,        // LP fees
        "l1_data_fees_usd": null        // populated on L2 rollups only
    },

    "cost_to_trade_24h": {
        "average_cost_per_trade_usd": 1.78,
        "fee_to_volume_ratio": 0.00038,
        "gas_per_volume": 0.000186
    },

    // Per-field coverage map. 1 = fully populated; <1 = partial; null = absent.
    "data_availability": {
        "totalFees": 1, "baseFees": 1, "priorityFees": 1, "builderTips": 1,
        "poolFees": 1, "l1DataFees": 0, "feeRegimeClassification": 1,
        "mevRiskLevel": 1, "mevToTotalFeesRatio": 1, "sandwichRate": 1,
        "gasPerVolume": 1, "feeToVolumeRatio": 1,
        "volume": 1, "liquidity": 1, "transactions": 1
    },

    "bars_hourly": [ /* 25 hourly bars with full MEV/fee/OHLCV columns */ ],
    "safer_trading_windows_utc": [ /* timestamps where MEV risk was 'low' */ ],
    "ai_verdict": "Medium MEV, 24% of fees are MEV, avoid this pool unless using a private mempool.",
    "fetched_at": "2026-04-30T13:00:07.235Z"
}

Reading data_availability

Codex doesn't index every chain or pool with the same depth. Field availability varies, especially on Solana where Jito-bundle data is per-pool. The map shows what populated for this row, so you can distinguish "no MEV detected" (zero) from "data unavailable" (null). Use this to filter or weight your analysis.

safer_trading_windows_utc

Hour-of-day timestamps (UTC) where the bar's mev_risk_level was "low" over the last 24h. Empty array on permanently-MEV-dominated pools.

Pricing

This actor uses pay-per-event pricing.

Event	Title	When it fires	Price
apify-actor-start	Actor start	Once per run (auto-fired by Apify, per GB memory)	$0.02
pair-scanned	Pair scanned (primary)	Per DEX pair returned with full intel + 24h hourly bars	$0.05
ai-verdict	AI verdict	Per Gemini-generated trader verdict (only when enable_ai_verdict=true)	$0.02

Worked examples:

• Scan one specific pair (no AI): $0.07
• Scan 5 pools of a token with AI: $0.02 + 5×$0.05 + 5×$0.02 = $0.37
• Discover top-20 MEV-toxic pools across 3 chains with AI: $0.02 + 20×$0.05 + 20×$0.02 = $1.42

Limits and edge cases

• Hard cap of 20 pairs per run (max_results setting). Designed for fast, focused scans rather than bulk dumps. Use scheduling for higher volume.
• Per-pool field availability varies on Solana — some Raydium/Orca pools return only sandwichRate, others return everything. Codex's Jito-bundle indexing isn't uniform across all pools. The data_availability map flags this per row.
• Builder tips are zero on Base / Arbitrum / Optimism / Polygon / other L2s — these chains don't have an MEV-builder market like Ethereum mainnet, so builderTips is genuinely zero (not missing data). mevToTotalFeesRatio will then also be null because the math has no numerator.
• l1DataFees populates only on L2 rollups (Base, Arbitrum, Optimism, etc.). On L1 chains it's null by design.
• Pair indexing churn — DEX pools rotate and are sometimes deprecated. If you scan an inactive pool address, you'll get an empty result with a warning. Use scan-token mode if you want the actor to discover currently-active pools.
• Free-tier safety on Solana addresses — the actor preserves case for Solana base58 addresses. Lowercasing breaks them. Just paste the address as-is.

Use cases

• Defensive trading — before placing a large swap, scan the pool and check mev_to_total_fees_ratio_24h. If high, route through a private mempool (Flashbots Protect, MEV Blocker).
• DAO treasury management — scheduled daily run on discover-mev for the chains your treasury is exposed to. Alert on new high-MEV pools where you have liquidity.
• Market-maker pool selection — when picking which DEX pool to LP into, prefer pool-fee-dominated regime over mev-dominated. The fee breakdown shows which pools genuinely return fees to LPs vs. leak them to MEV builders.
• MEV research / academic — bulk export hourly bars across many pools to study MEV intensity over time, validator dependence, etc.
• Trader copy-bot defense — before copying a trade signal, check if the source pool has historical sandwich activity.

Powered by

Codex.io — the GraphQL data API for 133 blockchain networks. This actor is a thin orchestration layer on top: it handles input validation, multi-pair fan-out, output normalization, AI enrichment, and Apify-native dataset/scheduling integration. All on-chain data sourced via Codex.

Limitations vs. our roadmap

What this v1 doesn't do (yet):

• Wallet-level smart-money tracking — Codex's wallet queries (filterWallets, detailedWalletStats) require their Growth plan. Roadmap item once we upgrade.
• Real-time webhook alerts — currently scheduled-poll-based via Apify cron. Native Codex webhooks are a v2 feature.
• Pre-trade slippage simulation — getting a quote for a hypothetical trade size and seeing how MEV/slippage scales is a v1.x candidate.
• Cross-pair "safer hours" consensus — when multiple pools are scanned, computing a chain-wide low-MEV window. v1.x.

Changelog

• 0.1.11 — Added default_evm_chain dropdown to the input form. Bare 0x addresses now resolve to the dropdown's chain (defaults to ethereum) instead of always defaulting to ethereum. Pick base once and paste bare Base addresses for the whole run.
• 0.1.10 — Auto-detect chain when address shape is unambiguous: bare 0x[40 hex] → ethereum, bare Solana base58 → solana. Helpful warnings when format is unparseable.
• 0.1.9 — Cleaner section names ("Choose what to scan", "Tokens — for the 'Scan a token's top pools' mode", etc.) replacing mechanical "Step 2A/B/C". Added items: { type: "string" } for proper stringList rendering.
• 0.1.8 — Switched pair/token inputs from raw JSON-array editor to stringList with chain:address strings. Backward-compat parser still accepts legacy {address, chain} object format.
• 0.1.7 — Store-ready README.
• 0.1.6 — Removed redundant manual actor-start charge; Apify auto-fires apify-actor-start per GB memory.
• 0.1.5 — Solana base58 addresses now case-preserved (EVM hex still lowercased). Fixes empty-result bug on scan-token for Solana / Sui / Aptos tokens.
• 0.1.4 — Disabled Gemini's default thinking mode (thinkingBudget: 0) and raised output budget to 120 tokens. AI verdicts now consistently produce full trader-language sentences instead of two-word labels.
• 0.1.3 — Logo + dataset_schema views (overview + fee breakdown).
• 0.1.2 — Initial public build.