Solana sandwich bots took $29.7 million before costs, mostly targeting Pump.fun
Overview
Over an 11‑month span, sandwich bots on Solana extracted about $29.7 million before costs, according to a study by Bitquery. The attacks hit more than six million trades and focused heavily on Pump.fun’s launch curve and PumpSwap.
The investigation covered 6,850,144 sandwich attacks by 2,545 bot wallets between Oct. 30, 2025 and Oct. 7, 2026. It found 96 % of the detected attacks targeted Pump.fun’s launch curve and PumpSwap, while other bots accounted for 39 % of the gross take.
Bitquery’s data also showed that trades with a 20‑50 % slippage tolerance were sandwiched 14 times as often as those with under 3 % tolerance.
Key Points
- Bitquery estimates $29.7 million in gross extraction from Solana sandwich trades over 11 months, before costs.
- Pump.fun’s launch curve and PumpSwap accounted for 96 % of detected attacks; other bots’ trades contributed 39 % of gross proceeds.
- Within observed Jupiter routes through those venues, 20‑50 % slippage trades were sandwiched 14 times as often as under‑3 % trades.
What Bitquery’s Study Says
The Bitquery investigation identified 6,850,144 sandwich attacks across 2,545 bot wallets, affecting 2,154,994 distinct wallets. The attacks were measured by repeated trading patterns; round‑trip trades beginning within a token’s first 10 slots were excluded. The study also notes gaps in exchange and block coverage, meaning the full scope may be larger.
Slippage tolerance limits set how far execution can depart from a quote. The data shows that higher slippage limits are more frequently exploited, with the 20‑50 % band being the most targeted.
Confirmed Findings
• Gross extraction before costs: about $29.7 million (valued at the minute of each attack).
• Attack period: Oct. 30, 2025 – Oct. 7, 2026.
• Bot count: 2,545 wallets performed 6,850,144 attacks.
• Affected wallets: 2,154,994 distinct wallets.
• Target venues: 96 % of attacks hit Pump.fun launch curve and PumpSwap.
• Other bots contributed 39 % of gross proceeds.
• Slippage exploitation: 20‑50 % slippage trades sandwiched 14x more often than under‑3 % trades.
• Fee recording inconsistency may affect the gross estimate by roughly 3 %.
• A separate 63‑attack fee sample showed bots retained 68 % of gross proceeds, while 18 attacks lost money after costs.
Uncertainties Remain
Chain records cannot prove that a bot saw a victim’s order beforehand. Some wallets that qualified as bots may actually be fast traders acting independently. Pump.fun’s sale‑fee logging is inconsistent, which impacts the accuracy of the gross estimate. Coverage gaps in exchanges and blocks also mean the true number of attacks could differ.
Why This Matters
Sandwich attacks reduce trader profits on Solana, especially on memecoin launch platforms like Pump.fun. When users see their trades executed at worse prices, trust in the ecosystem can decline, potentially discouraging participation and liquidity provision. The high concentration of attacks on a single launch curve also raises questions about the security of popular Solana DEX venues.
What Happens Next
Bitquery notes that the investigation is still evolving, and further analysis is planned to close coverage gaps. The findings are published as a research report, and interested readers can review the full study for methodology details. No immediate regulatory or platform actions are announced in the source material.
Sources
Newisty Editorial Team covers technology, cryptocurrency, digital products, online platforms, developer tools and the wider digital economy. Our content is researched from official sources, company announcements, public documentation, market data and other primary or reputable sources. Articles are reviewed and edited before publication for clarity, accuracy and useful context.
Related Posts
Polymarket Appoints Former Amazon Executive Warren Jenson as Its First CFO
