Solana Sandwich Bots: $29.7 Million, Mostly Pump.fun?

Updated 5 November 2026 · Standard formula, shown below

Yes—one widely reported estimate put Solana sandwich bots’ gross revenue at about $29.7 million over the period studied, before operating costs, and found that much of the activity targeted trades involving Pump.fun tokens. Treat both figures as study-specific estimates, not a verified total for every bot or a measure of net profit: the result depends on the time window, attribution method, and definition of revenue.

What the $29.7 million figure means

A sandwich bot watches for a pending trade, places a transaction ahead of it, and then attempts to trade in the opposite direction after it. The trades can move the market price against the original trader, allowing the bot to capture a spread. The reported $29.7 million figure refers to estimated proceeds from this activity in a particular analysis; it does not mean every Solana user lost that amount, nor that bots kept the full sum as profit.

“Before costs” matters. A bot may pay transaction fees, priority fees or tips to get transactions included quickly, and it can lose money when prices move unexpectedly or its transactions fail. A gross-revenue estimate does not automatically subtract those expenses, account for failed strategies, or establish how much was ultimately earned.

The figure should also be read with its study boundaries in mind. Different analyses can cover different dates, identify bots using different signals, and classify transactions differently. Unless a report clearly specifies its methods and period, its headline number should not be presented as a timeless or definitive network-wide tally.

Why Pump.fun comes up

Pump.fun is a token-launch platform associated with frequent trading in newly created tokens. New launches often have volatile prices, thin or changing liquidity, and bursts of attention. Those conditions can create opportunities for transaction-ordering strategies, which helps explain why analyses of Solana sandwich activity may find a large share connected to Pump.fun-related tokens.

That does not establish that Pump.fun itself runs sandwich bots, directs them, or is responsible for every sandwich involving a token launched there. A token’s association with a launch platform is not proof of who operates a bot. Nor does “mostly targeting Pump.fun” necessarily mean most Pump.fun trades are sandwiched: it may instead mean that Pump.fun-related activity accounted for a large share of the sandwiches identified by a particular study.

In short, the safest reading is that one analysis attributed a substantial portion of detected activity to trading around Pump.fun tokens. The exact share depends on what the researchers counted as a target and how they linked token activity to the platform.

How a sandwich attack works on Solana

In a conventional example, a user submits a swap that would buy a token. A bot sees the pending transaction and tries to buy immediately before it, raising the price. The user’s swap then executes at a worse price, within the limits the user allowed. The bot sells after the user’s transaction, seeking to collect the difference.

Spot a trade: The bot monitors transactions or other signals for a potentially profitable swap.
Compete for transaction order: It submits a trade designed to land before the target transaction, sometimes paying for faster or more reliable inclusion.
Let the target trade execute: The target’s purchase or sale can move the price in the bot’s favor.
Trade back: The bot submits the opposite trade, aiming to realize a gain after accounting for fees and price movement.

Solana’s high throughput and fast block production do not eliminate transaction-ordering competition. Searchers and validators may compete over transaction placement and inclusion, and the details vary with network conditions and trading infrastructure. A sandwich is therefore not simply “a fast trade”; it depends on the sequence of transactions and the effect that sequence has on the target’s execution.

Gross revenue is not net profit

Headline revenue numbers can sound like guaranteed earnings, but this strategy has costs and risks. A bot may pay fees and tips on both sides of a trade, fail to land one leg, or end up holding a token whose price moves against it. Competition can also reduce the available spread: several bots chasing the same opportunity may bid up execution costs or leave little profit after costs.

For a simplified example, suppose a bot appears to make 1 SOL on a completed sequence. If its transactions and execution costs total 0.2 SOL, its result before other expenses would be 0.8 SOL. If a leg fails or the token price changes before the bot can exit, the outcome could be smaller or even negative. This illustration explains the distinction; it is not an estimate of actual bot costs or returns.

Researchers may also have to infer which addresses belong to the same operator and distinguish sandwiching from other forms of arbitrage or market-making. An estimate can therefore be affected by attribution and classification choices, in addition to the usual question of which costs have been deducted.

What traders can do to reduce exposure

No setting guarantees that a swap will avoid adverse execution, especially in a fast-moving, low-liquidity token. Practical steps can reduce the room a sandwich strategy has to worsen a trade:

Check slippage tolerance: A wider limit can allow a swap to complete despite a larger price change. Use only a limit you understand and are willing to accept.
Review the quote and price impact: A large price impact is a warning that the trade itself may move the market substantially.
Consider trade size and liquidity: A smaller trade relative to available liquidity generally has less capacity to shift the price, though it does not remove all execution risk.
Be cautious with new-token launches: Thin liquidity and rapid price swings make outcomes less predictable. A token’s presence on a launch platform does not establish its quality or safety.
Inspect transaction details: Confirm the token, amount, route, and maximum acceptable execution conditions before signing.

These are risk-management measures, not a way to identify every bot or guarantee a favorable price. A transaction can receive poor execution for reasons other than a sandwich, including ordinary volatility, low liquidity, or a sudden change in demand.

Frequently Asked Questions

Did Solana sandwich bots make $29.7 million in profit?
The reported figure is a gross-revenue estimate before costs, not necessarily net profit. Fees, tips, failed transactions, and losing trades can change the final result.

Were most Solana sandwich attacks on Pump.fun?
One reported analysis found that much of the activity it identified involved Pump.fun-related tokens. That is not the same as proving that most Pump.fun trades were attacked, and the result depends on the study’s definitions and time period.

Is Pump.fun responsible for sandwich bots?
A bot’s activity around a token associated with Pump.fun does not show that the platform operates or controls the bot. Attribution requires evidence about the bot operator, not just the token’s origin.

Are sandwich bots the same as arbitrage bots?
They are related forms of automated trading but are not identical. Arbitrage typically seeks a price difference between markets or pools; a sandwich strategy deliberately places trades around a target transaction to benefit from its price impact.

Can I completely prevent a Solana swap from being sandwiched?
No general setting guarantees that. Sensible slippage limits, attention to liquidity and price impact, and careful review of transaction details can reduce some risks, but market movement and execution conditions remain uncertain.

Method: the standard formula for this calculation, rounded as each result states. See how the calculators work or report an error.