Solana Arbitrage Bot

High-performance arbitrage engine for Solana that captures price discrepancies across multiple DEX venues — Pump.fun, Meteora DLMM, and dAMM v2 — with a gRPC-driven market data and control plane.

Test transaction (demo run):
https://solscan.io/tx/3oUmy9CMdc2diGxhrkHcVUuWUBZnr7m862J8QxNELMy943T325wspgZmzHhbZsXRqPhEU68UZ3qpCuNSrHoSMTGC

Overview

This project is an on-chain arbitrage bot that:

• Monitors multiple Solana venues for mispricings in near real-time
• Computes optimal cross-venue routes (2-leg and 3-leg) including fees, slippage, and gas
• Submits atomic transactions to capture spread with strict risk controls
• Uses a gRPC interface for fast data ingestion, control, and observability

The system is designed for low-latency discovery and execution while remaining configurable and safe for continuous operation.

Key Features

• Cross-venue arbitrage: Pump.fun, Meteora DLMM, dAMM v2
• gRPC control plane: start/stop strategies, parameter updates, health checks
• Deterministic pricing with venue-specific fee and slippage modeling
• Atomic swaps with route validation and pre-flight simulation (when enabled)
• Configurable risk limits (max notional, per-venue caps, daily loss limits)
• Rate limiting and backoff to avoid RPC bans
• Structured logging and run-time metrics

Architecture

• gRPC Server: Accepts commands, streams metrics, and publishes market snapshots.
• Strategy Engine: Detects spreads, sizes trades, and enforces risk.
• Router: Selects the best executable path across venue adapters.
• Venue Adapters: Normalized interfaces for Pump.fun, Meteora DLMM, and dAMM v2.
• TX Builder: Crafts atomic transactions with compute budget tuning and prioritization fees.

How It Works

1. Subscribe to quotes/orderbooks via gRPC data channels and/or venue polling.
2. Normalize quotes to a common representation (price, size, fees, slippage curve).
3. Identify profitable cycles (2-leg or 3-leg) that exceed configured thresholds.
4. Simulate (optional) and validate all legs with current slots and account states.
5. Construct and send an atomic transaction; confirm with desired commitment.
6. Record results, update risk metrics, and continue scanning.

Requirements

• Node.js or Rust toolchain (depending on your implementation)
• Solana CLI and access to mainnet RPC + WS endpoints
• A funded keypair with enough SOL for compute/prioritization fees
• Optional: Prometheus/Grafana for metrics, Loki for logs

Configuration

Environment variables (example):

RPC_ENDPOINT=https://mainnet.helius-rpc.com/?api-key=
RPC_WEBSOCKET_ENDPOINT=wss://mainnet.helius-rpc.com/?api-key=
SLIPPAGE=10       # percent
PROFIT_LEVEL=10   # percent
COMMITMENT=processed
JITO_FEE=0.0003

GEYSER_RPC=wss://grpc

Per-venue settings can be provided via config files or environment variables (e.g., pool allowlists/denylists, token decimals, slippage curves).

Setup

# 1) Install dependencies
npm install

# 2) Create a .env file
cp .env.example .env
edit .env

# 3) Verify Solana CLI
solana --version
solana config get

# 4) Fund your keypair
solana balance

Running

# Start the gRPC server and strategy engine
npm run start

# Or run with explicit env overrides
GRPC_PORT=50052 ENABLE_PUMPFUN=false npm run start

gRPC interface (examples):

• Start/stop strategy
• Update thresholds (min spread, max notional)
• Stream metrics (PnL, win rate, latency)
• Subscribe to market snapshots

Strategy Controls

• Min spread threshold: Reject routes below net-profit threshold after all fees
• Slippage guard: Size trades to worst-case slippage per venue curve
• Health checks: Halt on repeated failed confirms, RPC errors, or staleness
• Cooldowns: Adaptive backoff after errors or consecutive reverts

Observability

• Structured logs with per-trade breakdown (route, expected vs realized PnL)
• Exported metrics: trades, fills, PnL, error rates, compute, confirmation latency
• Optional tracing for hot paths (routing, account fetching, simulation)

Test Results

• Example mainnet transaction: https://solscan.io/tx/3oUmy9CMdc2diGxhrkHcVUuWUBZnr7m862J8QxNELMy943T325wspgZmzHhbZsXRqPhEU68UZ3qpCuNSrHoSMTGC

Safety Notes

• Mainnet trading is risky; spreads can vanish before confirmation
• Always start with small size; validate configs and simulate
• Maintain sufficient SOL for compute/priority fees and rent
• Understand each venue's constraints and failure modes

Roadmap

• Add more venues and RFQ endpoints
• Improve path search (multi-hop, multi-venue) and sizing
• Portfolio-aware risk (inventory targeting, token caps)
• Enhanced strategy tuning via gRPC