LetsBonk sniper bot: sniping Bonk.fun on LaunchLab rails

Why a LetsBonk sniper bot is really a LaunchLab bot

LetsBonk did not build its own bonding curve. It launched on top of Raydium's LaunchLab infrastructure, which means the instructions your bot has to decode are LaunchLab's instructions and the accounts it has to read are LaunchLab's accounts. A bot written against Pump.fun sees nothing here, because it is listening for a program that is never called.

This is the single most common reason a sniper bot claims LetsBonk coverage and still misses launches. Claiming a launchpad is a marketing line. Decoding its program is engineering, and the two are not the same.

What changes when the rails are shared

• The program ID your listener subscribes to belongs to Raydium, so every LaunchLab launch arrives in the same stream, LetsBonk or not.
• Curve parameters are configuration rather than constants, so you cannot hardcode a curve shape the way you can on a venue that ships one fixed formula.
• The graduation path is Raydium's own AMM, which is the venue your exit logic already knows, so the post-graduation half of the trade is unusually familiar.

Telling a Bonk.fun launch apart from every other LaunchLab launch

Because LetsBonk runs on shared rails, filtering is not optional. Raydium's LaunchLab serves its own launches and other front ends as well, so a listener that simply watches the program and buys will occasionally buy something that has nothing to do with Bonk.fun and none of the flow you were counting on.

The separation lives in the platform configuration attached to each launch. LaunchLab lets a platform register its own config, carrying its fee split and its LP disposition, and every launch created through that front end points at it. Read that account and you know whose launch you are looking at before you size anything.

Quote mint deserves its own line. A SOL denominated curve graduates around 85 SOL; a USDC denominated one graduates near 58,783 dollars. A bot that assumes SOL and meets a USDC launch will misjudge how far along the curve is by a wide margin, and it will size accordingly.

What an 85 SOL graduation does to the trade

Graduation thresholds are not trivia. They decide how much buying has to arrive before the curve ends, and therefore how long the curve phase lasts and how much room there is between your entry and the moment everybody else can reach the token.

Eighty five SOL is a low bar. A launch with any real attention behind it can clear it quickly, which means the window where you are early is narrow and the crowd that waits for a Raydium pool is never far behind. On a venue with a distant threshold you can hold through the curve and think about it. Here the curve is a corridor, not a room.

The practical consequence is that a LetsBonk plan needs its exit decided before entry. If your thesis is that graduation itself is the liquidity event you are selling into, then you need the position sized so it can actually leave through an 85 SOL curve without eating your own gain on the way out.

Where the 1% goes, and why a sniper should care

LetsBonk charges a 1% trading fee on curve activity. Published breakdowns route it across RAY buybacks, platform operations and creator incentives, with a portion of platform revenue funding BONK buybacks and burns. That is the ecosystem story, and it is genuinely different from a launchpad that simply keeps the fee.

For your purposes the destination matters less than the arithmetic. One percent is charged on the way in and again on the way out, so a round trip starts a full two percent behind before slippage, priority fees and the spread you cross at graduation. On a trade you expect to hold for ninety seconds, that is not a rounding error.

That fourth row is where most of the quiet losses live. Every rejected attempt still costs the priority bid you attached to it, so a bot that retries aggressively into a curve it cannot fill is spending money to lose races.

What to screen when the contract is boilerplate

The same trap that catches people on Pump.fun catches them here. Launchpad tokens are minted from a template, so the checks most screeners lead with are answered identically on every single launch. If mint authority is revoked by design, seeing it revoked tells you the launchpad works. It tells you nothing about the token.

What actually separates launches on a shared, templated venue is human. Who deployed it, what they deployed before, how the opening buys were arranged and how much of the supply moved before anybody outside could reach it.

1. 1Read the creator, not the contractPull the deployer's history. A wallet with a trail of launches that died at the same point in the curve is telling you exactly what this one is for.
2. 2Look at the opening slotIf a large share of supply left the curve in the first slots across a cluster of wallets, you are not early. You are the counterparty for people who were.
3. 3Follow the funding graphWallets funded minutes earlier from one source, all buying together, is a bundle wearing a disguise. It is visible before you buy if you look.
4. 4Check distance to graduationOn an 85 SOL curve, knowing whether 10 SOL or 70 SOL has already arrived changes the trade completely. It is the single most useful number on the screen.
5. 5Confirm the quote mintSOL or USDC changes every threshold calculation you are about to make. Confirm it rather than assuming it.

Graduation, the LP burn and the Jupiter moment

When the curve completes, liquidity migrates into a Raydium pool and the LP tokens are burned. Burned LP is worth understanding precisely: it removes the classic rug where a deployer pulls liquidity out from under holders, because there is no longer an LP position anybody can withdraw.

It does not remove risk. It removes one mechanism. Supply concentrated in a handful of wallets can still be sold into you, and on a freshly graduated pool that sale has somewhere to land for the first time.

What changes the second the pool exists

• The token becomes routable, so aggregation picks it up and buyers who never touch launchpads can reach it without doing anything special.
• Depth arrives, which means the slippage you needed on the curve is now far too wide and will cost you real money on every fill.
• The trade changes character from a race into an ordinary pool trade, and the settings that won the race are the wrong settings for what follows.

What LetsBonk snipers get wrong

1. 1Pointing a Pump.fun listener at itDifferent program, different instructions, different accounts. The bot does not error. It simply never sees a launch and you assume the venue is quiet.
2. 2Skipping the platform filterLaunchLab carries launches from more than one front end. Without reading the platform config you are buying whatever the program emits.
3. 3Assuming a fixed curveCurve parameters are configuration here. A bot that hardcodes one shape will misprice fills on launches configured differently.
4. 4Treating 85 SOL like 400 SOLThe curve phase is short. A plan that assumes time to watch, average in and decide is a plan for a different venue.
5. 5Reading a burned LP as safetyIt closes one door. Concentrated supply walks through a different one, and it does so after graduation, not before.

Questions people ask

These come up in the ticket queue often enough to answer once, properly. If yours is not here, send it over and it probably ends up on this page.

Key takeaways

• LetsBonk runs on Raydium LaunchLab, so sniping it means decoding LaunchLab instructions rather than Pump.fun ones.
• LaunchLab carries launches from more than one front end, so the platform config is the real Bonk.fun filter.
• Graduation sits near 85 SOL on SOL pairs and near $58,783 on USDC pairs, and the quote mint decides which applies.
• A low threshold shortens the curve phase, which makes a pre-decided exit more important than a fast entry.
• The 1% fee is charged both directions, so a round trip starts two percent behind before slippage and tips.
• LP is burned at graduation, closing the liquidity pull rug and leaving concentrated supply as the risk that remains.