Stream Raydium swaps from the validator.

Raydium is Solana's largest decentralized exchange by trading volume. It operates three distinct trading mechanisms: AMM V4 for standard constant-product swaps, CLMM (Concentrated Liquidity Market Maker) for capital-efficient concentrated liquidity positions, and the newer Raydium Launchpad for token launches and initial liquidity bootstrapping. Each mechanism serves a different market segment, and each generates a massive volume of on-chain transactions that traders need to monitor.

The complexity starts with Raydium's multi-program architecture. AMM V4 runs under the program ID 675kPX9MHTjS2zt1qfr1NYHuzeLXfQM9H24wFSUt1Mp8, while the CLMM concentrated liquidity pools operate under CAMMCzo5YL8w4VFF8KVHrK22GGUsp5VTaW7grrKgrWqK. These are separate programs with different instruction layouts, different account structures, and different event formats. Monitoring Raydium comprehensively means subscribing to both programs and handling their distinct data schemas.

Each of these programs processes thousands of transactions per minute during active trading periods. During token launches or market-moving events, throughput spikes dramatically as bots and traders compete for execution. The sheer volume makes traditional RPC polling impractical — by the time your bot polls for new transactions, parses the response, and identifies a relevant event, the opportunity window has already closed.

Different trading strategies require different Raydium events. For price feed and arbitrage strategies, you need every swap event in real-time — each swap shifts the pool's price curve, and your bot needs to reflect that change immediately to detect cross-DEX discrepancies. For sniping and launch trading, you need pool creation events before anyone else sees them — the first few seconds of a new pool's life produce the widest spreads and the largest profits. For liquidation bots and risk management systems, you need liquidity removal events instantly — a large withdrawal from a pool changes its depth and price impact characteristics, and positions that were safe at one liquidity level may become vulnerable at another.

Every Raydium swap transaction contains a set of critical fields that your bot needs to extract and process. The core data points are: the pool address identifying which liquidity pool the swap occurred in, the token_in object containing the input token's mint address and the raw amount deposited, the token_out object with the output token's mint and the amount received, the direction indicating whether the trade was a buy or sell relative to the base token, and the price_impact measured in basis points showing how much the swap moved the pool's price.

Calculating the effective price from a swap event is straightforward once you have the parsed fields. Divide the token_out amount by the token_in amount, adjusting for each token's decimals. For example, if a swap sends 1,500,000,000 lamports of SOL (1.5 SOL at 9 decimals) and receives 42,819,200 units of a 6-decimal token, the effective price is 42,819,200 / 10^6 divided by 1,500,000,000 / 10^9 — which gives you 28.546 tokens per SOL. Tracking this price across consecutive swaps in the same pool gives you a real-time price feed with slot-level granularity.

One of the most practical applications of a Raydium gRPC stream is building a real-time price feed. Every swap that flows through a Raydium pool shifts the pool's reserves and therefore its effective price. By processing these swap events as they arrive, you can construct a price ticker that updates with every trade — not every few seconds like a polling-based approach, but with every single on-chain swap.

The calculation is straightforward. For each swap event, divide the output amount by the input amount (adjusted for decimals) to get the effective exchange rate. Store this alongside a timestamp derived from the slot number, and you have a tick-by-tick price history. Aggregate these ticks over time windows — 1 second, 5 seconds, 1 minute — and you have volume metrics for each pool. Track the cumulative token_in amounts for volume, and count the number of swap events for trade frequency.

This data feeds multiple downstream use cases. For arbitrage detection, compare the Raydium price against other DEXs — any discrepancy beyond your gas + slippage threshold is an opportunity. For portfolio trackers, map token holdings to real-time prices without relying on third-party oracle services. For dashboards, display live trading activity with per-pool granularity. For on-chain risk systems, monitor price impact trends to detect liquidity drain or manipulation attempts.