The trade took less than a second. It cost 2.32 SOL in fees. It returned $696,000 in USDC. The bot that executed it did not predict the market. It simply noticed that the same token was trading at two different prices simultaneously and closed the gap before anyone else could.
This is not a hack. This is not a bug. This is Solana working exactly as designed.
What Happened
ANB, the token for Ant Blockchain, had two liquidity pools running simultaneously on Meteora. One was a Dynamic AMM pool. The other was a Dynamic Liquidity Market Maker pool. Both listed ANB. Both had prices. And for a brief window yesterday, those prices were catastrophically different.
A single large sell order caused 99% price impact in one pool, crashing ANB to near zero while the parallel pool continued trading at the prior price. The mismatch was instant and enormous. Buy ANB cheap in the crashed pool. Sell at the prior price in the stable pool. The entire sequence executes in one atomic transaction. If either leg fails the trade reverts with no loss.
One wallet turned $0.10 into $196,000. Another converted $0.036 into $86,714. The largest single extraction was $696,000 from a 2.32 SOL input. Total profits across two consecutive blocks reached approximately $1.32 million before the pools rebalanced.
13 hours ago, multiple wallets ran arbitrage trades on $ANB
They turned less than $1 into $60K–$696K+ in a single trade. Multiple times.$ANB went from ~$67M MC down to ~$30M MC.
How it worked:$ANB had two pools on Meteora at the same time:
1. A Meteora DAMM v2 pool where… https://t.co/R0bhzqCOD3 pic.twitter.com/RMBWPC9Bmx
— Kakashi (@kkashi_yt) May 2, 2026
How the Winning Bot Did It
The specific route went like this. A tiny USDC position was swapped for roughly 72,793 ANX tokens on a Meteora DAMM v2 pool. Those ANX tokens were immediately routed into a second pool where they purchased more than 50 million ANB at the crashed price. The entire ANB position was then sold through Jupiter Aggregator v6 at the pre-crash price. The round trip yielded 885,290 USDC before fees. Total time elapsed: a fraction of a second.
The infrastructure that made this possible is Jito, Solana’s MEV bundle system. Jito allows bots to submit transaction bundles that execute atomically across multiple steps. No partial fills. No stuck positions. The bots that extracted $1.32 million were not just fast. They were architecturally closer to the price data than any human trader could be.
The Part Nobody Wants to Say
The bots did nothing illegal. They identified a price mismatch the market created and closed it faster than any human could.
But every dollar extracted came from liquidity pools funded by liquidity providers who did not knowingly accept the risk that a single large trade could create a $696,000 extraction event in under a second. ANB dropped from a $67 million market cap to approximately $30 million. Retail holders watched 99% of price evaporate before they could react.
The dual-pool setup that created this vulnerability is not unusual on Solana. What was unusual was the size of the sell order that opened the gap. But sell order size is not predictable or controllable.
The retail trader who held ANB did not lose money to a hack. They lost money to a market structure where bots are architecturally positioned to extract value from any price dislocation faster than any human can respond.
That is not a bug. That is the current state of decentralized trading.
