Solana sandwich bots extracted $29.7 million in gross revenue before costs over an 11-month period between October 30, 2025, and October 7, 2026, according to an investigation by blockchain data provider Bitquery.
Heavy Exploitation Centered on Pump.fun
Bitquery identified 6,850,144 attacks executed by 2,545 bot wallets, impacting 2,154,994 distinct addresses. A overwhelming 96% of all detected sandwich attacks occurred on memecoin platform Pump.fun's launch curve and its PumpSwap exchange. Notably, trades initiated by other automated bots accounted for 39% of the estimated gross proceeds.
Sandwich attacks execute by placing a bot buy order before a targeted swap and a sell order immediately after. The initial purchase inflates pool prices, leaving the victim trader with a worse execution price while the bot profits from selling at the elevated level.
Slippage Limits and Cost Dynamics
For trades routed by Jupiter through Pump.fun or PumpSwap, transactions allowing 20%–50% slippage were sandwiched 14 times as frequently as those allowing under 3%. However, attack frequency did not scale linearly across all bands: trades under 0.5% slippage tolerance were attacked more often than those at 1%–3%, whereas settings of 50% or more experienced fewer attacks than the 20%–50% range. Algorithmic activity continues to dominate transaction dynamics across crypto protocols, similar to how trading bots reshape arbitrage execution windows.
The $29.7 million total values SOL at the exact minute of each trade and does not reflect net profits. In a granular sample of 63 attacks, bots kept 68% of aggregate gross proceeds, while 18 attacks incurred net losses after transaction fees. Bitquery also noted that inconsistent sale-fee recording on Pump.fun introduces a 3% margin of variance in its gross estimate.
Key Takeaways
- $29.7M Extracted: 2,545 bot wallets conducted 6.85 million sandwich attacks over 11 months.
- Pump.fun Targeted: 96% of all detected attacks occurred on Pump.fun's launch curve and PumpSwap.
- Slippage Exposure: Trades permitting 20%–50% slippage suffered a 14-fold increase in attack rate compared to trades under 3%.
Why It Matters
This research highlights how maximal extractable value (MEV) tactics exploit high-frequency retail trading venues on fast networks like Solana. While traders frequently widen slippage settings to guarantee execution during fast price moves, doing so leaves them vulnerable to extreme price extraction by front-running bots. Furthermore, as noted in a multi-chain research preprint published on September 23, 2026, existing protected order flow mechanisms across six blockchains often fail to offer the full protection users expect against cross-block sandwich patterns.
