MEV Sandwich Bot Development for Ethereum
We develop professional sandwich bots that exploit the slippage tolerance of DEX users — a proven MEV extraction strategy. Using Rust, Flashbots, and custom simulators, our bots execute arbitrage, liquidations, and sandwiches with over 95% inclusion rate. Specifically, a sandwich bot spots a victim in the mempool, inserts its own purchase before the victim's (frontrun), the victim moves the price up, and the bot sells after (backrun). The difference is the bot's profit, while the victim loses only within their slippage tolerance. This is arbitrage on transaction ordering information.
Developing such a tool is a challenge of speed (mempool latency), economics (profitability calculator), and infrastructure (reliable deployment without losses). With over 10 years in blockchain development and 50+ MEV bots deployed, we create profitable bots in highly competitive environments. We'll assess your project and provide a detailed profitability calculation.
Why Most Sandwich Bots Are Unprofitable
Inaccurate Profit Calculation
The most common mistake is calculating profit without accounting for all costs. Real P&L:
profit = amountOut_backrun - amountIn_frontrun costs = gasCost_frontrun + gasCost_backrun + builderTip netProfit = profit - costs gasCost with aggressive tips on Ethereum mainnet is not negligible — often $3–$8 per bundle. If a sandwich yields $10, but the builder tip is $5, net profit is only $5. The bot must simulate the entire bundle before submission and discard deals with negative net profit. Simulation via eth_call is slow. We use in-process EVM simulators (revm in Rust) for microsecond evaluation, achieving a median delay under 50 ms.
Revert Due to Price Impact Calculation
The bot calculated profit at time T, sent the transaction after 50 ms. In that interval, another bot executed its own frontrun. The pool state changed, price moves more than calculated, the victim exceeds their slippage tolerance, and their transaction reverts. The backrun is unnecessary, but gas for the frontrun is wasted — average cost per failed attempt: $2–$4. Protection: the backrun transaction is conditional — it verifies that the target transaction was actually included. In Flashbots bundles, this is realized via reverting_tx_hashes — if the victim reverts, the entire bundle is dropped.
Competition and Builder Relationships
On Ethereum, 95%+ of MEV bundles go through block builders (Flashbots, beaverbuild). A bot sending transactions directly to the public mempool loses. Proper infrastructure: MEV-Share, direct connections to top builders, monitoring of mev-boost relays. We ensure integration with leading relays to maximize inclusion rate, typically over 90%.
How to Optimize Execution via Flashbots?
The architecture of a production sandwich bot includes:
- Mempool monitoring with direct peer-to-peer connections and bloXroute feeds. We set up a custom node with
--txpool.pricelimit 0. - Candidate filtering: only known DEX routers, decoded calldata with slippage > 0.5%, sufficient gas price.
- Optimal frontrun amount calculation: for Uniswap V2 — analytical solution, for V3 — numerical optimization. Account for marginal cost of each additional dollar.
- Execution via Flashbots with conditional bundle:
const bundle = [ { signer: wallet, transaction: frontrunTx }, { hash: victimTxHash }, { signer: wallet, transaction: backrunTx } ]; const bundleResponse = await flashbotsProvider.sendBundle( bundle, targetBlockNumber, { minTimestamp: 0, maxTimestamp: 0, revertingTxHashes: [victimTxHash] } ); revertingTxHashes is key: if the victim reverts, the entire bundle is dropped, the bot doesn't lose gas on a frontrun without a backrun.
Chain Comparison for Sandwich Attacks
| Chain | Mempool Type | Competition | Avg Profit per Trade |
|---|---|---|---|
| Ethereum | Flashbots/private | High | $5–$50 |
| BSC | Public + private | Medium | $2–$20 |
| Arbitrum | Sequencer (limited) | Low | $1–$10 |
| Polygon | Public | Medium | $3–$30 |
Our bots operate on all these chains. For each, we select the optimal strategy — on Ethereum, focus on builder relationships; on BSC, speed and low fees. Average daily net profit per bot can range from $200 to $2000 depending on market volatility.
Key Efficiency Metrics
| Metric | Target Value |
|---|---|
| Inclusion rate | > 90% |
| Average profit per bundle | $8 |
| Revert rate | < 5% |
| Delay from tx receipt to bundle submission | < 50 ms |
We monitor every metric via Prometheus + Grafana. We guarantee transparency of operations.
What's Included
- Analysis of target chain and DEX ecosystem (2–3 days)
- Development and testing (2–3 weeks for base version)
- Deployment on mainnet with limited positions and optimization (1–2 weeks)
- Delivery of full documentation, infrastructure access, team training
- 1-month post-launch support
Work Process
- Research — assessment of MEV opportunities via Dune Analytics and custom scripts.
- Design — selection of stack, architecture, configuration of mempool feeds.
- Development — coding in Rust, smart contract deployment, Flashbots integration.
- Testing — backtesting on historical blocks in Foundry fork, paper trading on testnet.
- Launch — mainnet deployment, A/B parameter testing.
- Optimization — continuous tuning to current market conditions.
Estimated Timelines and Pricing
- Base Uniswap V2/V3 bot: 1–2 weeks — $5,000–$8,000
- Multi-DEX + multichain: 3–5 weeks — $12,000–$18,000
- Timelines include all stages from research to optimization
Pricing is project-based; we provide a fixed quote after assessment. Typical ROI: bot recovers development cost within 2–4 months of operation.
Tech Stack
We write the bot in Rust with the alloy library for maximum performance. Critical components — EVM simulator (revm), mempool streaming, bundle submission — are in a single process without inter-process communication. Smart contract for atomic execution is in Solidity 0.8.x with minimal footprint. Monitoring: Prometheus + Grafana.
Rust is 10x faster than TypeScript for transaction simulation — giving an edge in the MEV race.
Common Mistakes in Self-Development
- Using ready-made open-source bots — they are already exploited, competition is high.
- Ignoring gas costs for empty blocks — the bot can operate at a loss.
- Lack of revert conditions — gas loss on every failed frontrun.
We eliminate these issues at the design stage. Contact us for a free consultation — we'll assess profitability and propose the optimal solution.
Configuration Details
Bot parameters are tuned individually: minimum trade profit, maximum gas price, pool list; private RPC endpoints.Source: Flashbots Documentation MEV on Wikipedia Contact us for a free project assessment. Get a consultation on strategy and chain selection.