If you're deep into the world of meme tokens on Solana, you've probably felt the sting of a sandwich attack at some point. These sneaky maneuvers can turn a promising trade into a costly lesson. But there's good news on the horizon. The team behind Sandwiched.me, a Solana MEV dashboard, has rolled out Sandwich Detector V2, and early signs show it's making a real difference. Let's break it down.

Understanding Sandwich Attacks in the Meme Token Space

First off, what exactly is a sandwich attack? In blockchain terms, especially on networks like Solana where meme tokens thrive on platforms like Pump.fun or Raydium, a sandwich attack happens when a malicious actor spots your trade in the mempool (the queue of pending transactions). They then "sandwich" your transaction by placing one order before yours (frontrun) to buy the token, driving up the price, and another after (backrun) to sell at the inflated price, pocketing the difference. You end up paying more or getting less than expected due to slippage.

These attacks are particularly rampant in the volatile world of meme tokens, where quick pumps and dumps attract bots looking to extract value. MEV, or Maximal Extractable Value, refers to this profit that validators or searchers can grab by reordering transactions in a block.

As shown in the diagram above, modern attacks have gotten more sophisticated. Attackers might split their frontruns and backruns across multiple transactions or even transfer tokens to different wallets to evade detection. This cat-and-mouse game has made it tough for tools to keep up.

The Launch of Sandwich Detector V2

On August 28, 2025, Chris Chang, co-founder of Ghost Logs, announced the rollout of Sandwich Detector V2 on Sandwiched.me via this thread on X. The update aims to catch these advanced evasion tactics head-on.

Key features of V2 include:

• Processing data from over 50 Automated Market Makers (AMMs) on Solana.
• Analyzing swaps, token transfers, and other metadata to identify split or obfuscated sandwiches.
• Allowing users to toggle between V1 and V2 detection modes for validators' daily sandwich rates.

The team credits community efforts, including insights from @Shiroi_sol, for highlighting new strategies and pushing for reduced toxic MEV. This collaborative approach is crucial in a decentralized ecosystem like Solana.

Observed Trends Post-Launch

Just a few days later, on September 2, 2025, Chris shared an update in this tweet, noting intriguing trends. The charts reveal a noticeable decline in "sandwich blocks"—blocks containing detected sandwich attacks—following the V2 announcement.

Looking at the data from early August to September 2, sandwich blocks peaked around mid-August but have tapered off. Chris speculates that validators might be switching to new evasion methods or perhaps entering a "redemption era" by cleaning up their act. Either way, the increased scrutiny from V2 seems to be deterring malicious behavior.

A slightly updated chart from later that day shows similar patterns, reinforcing the trend.

What This Means for Meme Token Traders

For those hunting the next big meme coin, this development could mean fairer trades and less unexpected losses. With better detection, stake delegators—like those using Marinade Finance or Jito—can avoid supporting validators involved in toxic MEV, pressuring them to behave ethically.

If validators reduce sandwiching, meme token launches might see less bot interference, allowing organic growth and reducing the frustration of getting frontrun on that hot new token. It's a step toward a healthier Solana ecosystem, where retail traders aren't constantly outmaneuvered by sophisticated bots.

Of course, the battle isn't over. As Chris notes, detection is an ongoing game. But tools like Sandwich Detector V2 empower the community to monitor and respond.

If you're staking SOL, consider delegating to ethical validators like Ghost's own to support these efforts. Check out Sandwiched.me to see the data for yourself and stay informed on the latest in Solana MEV.

Stay vigilant, and happy memeing!