Pump.fun Sniper Bot — Snipe Solana Launches in Under 100ms

What Is a Pump.fun Sniper Bot?

Pump.fun remains the highest-volume token launchpad on Solana in 2026, with 8,000-15,000 new tokens launching daily. The platform's bonding curve mechanism prices tokens on a strict first-buyer-pays-least model — the first buy transaction after creation gets the minimum price, and every subsequent transaction pays more. In practice this means the first 3-5 seconds of a token's life determine whether you pay $0.00002 per token or $0.00008 per token. Sniping automates reaction speed, making it possible to secure those first slots across hundreds of launches per day without sitting in front of a screen.

The Pump.fun sniper bot market split in late 2024 between custodial web bots (BonkBot, Photon) and non-custodial Telegram-native bots (OpenLiquid, Trojan, Maestro, Banana Gun). Non-custodial has become the dominant pattern for sniping because speed advantages are marginal at this point while security concerns with custodial key storage are not — 2024 saw multiple custodial bot incidents where operator compromise led to user wallet drains.

How Pump.fun Sniping Works

Step 1: Detect new token mints

The sniper bot subscribes to Solana program logs for the Pump.fun program ID. When a new token is created, the program emits a log entry containing the mint address, deployer wallet, initial bonding curve state, and metadata reference. Detection happens via WebSocket subscription to a low-latency RPC endpoint — typically a dedicated RPC with sub-20ms geographic latency to Solana validators.

Step 2: Apply filters

Each detected launch is passed through the user's configured filter set in 1-5 milliseconds. Filters check on-chain state (deployer wallet history, initial buy sizes, token account flags) and off-chain metadata (social links, image availability, description content). Launches that fail any filter are dropped silently; launches that pass move to bundle construction.

Step 3: Build and sign Jito bundle

The bot constructs a transaction with the user's configured buy size, slippage tolerance, and priority fee. The transaction is bundled with a Jito tip transaction and signed using the user's encrypted wallet key (decrypted in-memory, never persisted). The Jito bundle guarantees atomic execution — either all transactions in the bundle land in the same block, or none do.

Step 4: Submit to Jito and confirm

The signed bundle is submitted to Jito's block engine via gRPC. Jito forwards the bundle to the current Solana block leader for inclusion. A successful bundle lands in roughly 400-800ms from submission; the bot then confirms the on-chain transaction state and notifies the user via Telegram with the buy price, token balance, and current bonding curve position.

Latency Budget: Why Milliseconds Matter

Pump.fun sniping is a latency-sensitive game because the bonding curve price increases monotonically with each buy. Every millisecond between token creation and your buy landing costs you a worse entry price and potentially your first-buyer slot entirely.

Achieving sub-100ms end-to-end requires dedicated infrastructure: a Solana RPC endpoint within 20ms network distance, WebSocket subscriptions (not polling), pre-built transaction templates to skip construction time, and Jito bundle tips calibrated to current block leader economics. OpenLiquid's sniper bot operates on dedicated RPC infrastructure with typical detection-to-landing times of 60-90ms, placing it in the competitive tier without requiring users to manage their own infrastructure.

Anti-Rug Filters: What the Sniper Screens For

Most Pump.fun launches fail. Some fail quickly (rug within 60 seconds of launch), some fail slowly (token fails to graduate and decays over hours). A sniper bot without filters is a machine for depositing SOL into immediate rugs. OpenLiquid's sniper bot exposes the following filters, all configurable per-session:

• Minimum dev buy size: Skip tokens where the creator's own initial buy is under a configurable SOL threshold (default 0.5 SOL). Sub-0.5-SOL dev buys correlate strongly with instant-rug attempts.
• Deployer wallet reputation: Cross-reference the deployer address against a maintained list of known rug wallets. Repeat ruggers are automatically blocked regardless of the new launch's appearance.
• Social metadata presence: Require Twitter, Telegram, or website links in the token metadata. Absence of any social presence is a strong rug signal.
• Name and image copy detection: Skip launches that appear to clone the name, ticker, or image of a currently-trending token. These are typically ride-the-wave scams targeting search misclicks.
• Minimum first-minute buyer count: After your sniper buy lands, require a minimum number of additional buyers within 60 seconds before the sniper considers the launch "real." If the first-minute buyer threshold isn't met, the sniper executes its configured exit rule automatically.
• Blacklist keywords: User-maintained word lists to auto-skip token names containing scam-associated terms or disallowed content.
• Creator allocation cap: Skip launches where the creator's post-bundle token holding exceeds a threshold (default 15%). High creator concentration precedes dump events.

Filter tuning is an active decision. Tight filters reduce false positives but miss more winners; loose filters catch more winners but also more rugs. Most experienced snipers run two profiles: a tight profile for capital preservation mode and a loose profile during high-signal market periods.

Getting Started in 60 Seconds

1. Open Telegram and start a chat with @OpenLiquidBot.
2. Create a sniper wallet. The bot generates a fresh Solana keypair encrypted with a passphrase only you know. Fund it with 1-5 SOL to start.
3. Configure snipe parameters. Set your per-snipe buy size (e.g. 0.1 SOL), slippage tolerance (default 15% — new launches move fast), and Jito priority tip (0.001-0.01 SOL depending on how aggressive you want to be).
4. Enable filters. Pick a preset (Conservative, Balanced, Aggressive) or configure individual filters. Conservative is recommended for your first session.
5. Set exit rules. Configure take-profit ladders (e.g. sell 30% at 2x, 40% at 5x, 30% trailing stop) and auto-sell on filter failure.
6. Start sniping. The bot enters monitoring mode. When a launch passes your filters, the bot executes the buy and notifies you with transaction details. You can pause, adjust, or stop at any time.

Sniper Strategy & Position Sizing

Sniper profitability is dominated by position sizing and exit discipline, not by raw hit rate. Most profitable snipers lose money on 60-80% of individual snipes and make their returns on the 20-40% that 5x or better.

Position sizing rules

• Never size a single snipe above 1-2% of your total sniper pool. With 10 SOL pool, that's 0.1-0.2 SOL per snipe.
• Increase size for passes of tighter filters. A launch that passes Aggressive filters gets base size; a launch that passes Conservative + has a known reputable deployer can get 2-3x size.
• Decrease size during macro weakness. In market-wide downtrends, most launches fail. Reduce per-snipe size by 50% until trend flips.

Exit discipline

• Take profit in layers. A standard ladder: 30% at 2x, 30% at 5x, 30% at 10x, 10% trailing stop. This captures outcome across the full win distribution.
• Accept quick exits. If a launch fails to attract buyers within 60-120 seconds, sell half immediately and let the rest run on trailing stop. Early traction is the strongest signal of whether a token will sustain.
• Never hold a snipe past graduation without a reason. Migration from bonding curve to PumpSwap changes token dynamics — re-evaluate the thesis before letting a position ride.

OpenLiquid vs Other Pump.fun Sniper Bots

The primary differentiator for OpenLiquid is the integrated tool suite: the same Telegram interface that handles sniping also provides rug checking, bundle launches, Solana volume bots, and multi-wallet management. For traders running multi-strategy workflows (snipe new launches, run volume bots on holdings, distribute winnings across cold wallets), the single-interface model eliminates the context-switching cost of jumping between 4-5 different bots.

Risks & Mitigation

Sniping Pump.fun launches is a high-variance activity. Understanding the risks is table stakes:

• Honeypot tokens: Contracts where sells are blocked or taxed at 99%. Mitigation: sell-simulation filter (the bot simulates a sell before confirming a buy as successful — if simulation fails, it flags the token).
• Instant rugs: Deployer dumps their allocation seconds after launch. Mitigation: creator-allocation filter (skip launches where deployer holds 15%+ post-bundle).
• Slow decay: Most tokens simply fail to attract buyers and decay 90%+ over hours. Mitigation: first-minute-buyer-count filter + quick exit on low traction.
• MEV-adjacent front-running: Unlike public mempool chains, Solana has less traditional MEV, but sniper-vs-sniper competition for the same launch is fierce. Mitigation: higher Jito tips during congestion, dedicated RPC, non-competing strategy (e.g. snipe after first-minute signal rather than at-mint).
• Platform risk: Pump.fun itself could change its bonding curve mechanics, add anti-bot measures, or be regulated. Diversify strategies across Pump.fun, PumpSwap, and Raydium launches rather than relying entirely on one platform.
• Key management risk: Your sniper wallet sits with SOL deposited and is active 24/7. Never put more than your defined "sniping pool" in this wallet — always keep the main treasury separate. Rotate the sniper wallet monthly as operational hygiene.

Frequently Asked Questions