Raydium

Raydium alternatives are shaped by Pool Depth and Swap Price Impact

Raydium alternatives are Solana trading venues and routers whose useful depth determines how far a swap moves the execution price. Orca Whirlpools concentrates liquidity inside chosen ranges, Meteora DLMM places it in discrete bins, Jupiter combines several sources and OpenBook V2 matches queued orders. The right comparison therefore starts with the same token mints, direction and input size, then checks quoted output rather than a protocol-wide liquidity total. This page focuses on that decision: where executable liquidity sits, how each mechanism consumes it and when a route exposes less price impact than a direct Raydium pool for the intended trade.

Headline TVL hides the depth that a swap actually reaches

The common mistake is treating total value locked as executable depth, although a swap touches only liquidity available along its exact price path.

A constant-product pool exposes both reserves continuously, while a concentrated-liquidity pool activates only positions covering the market price. Meteora DLMM adds another layer: inventory sits in separate bins, so nearby bins matter before distant capital. OpenBook V2 has no reserve curve; its relevant depth is the sum of limit orders available at successive prices. Protocol-wide liquidity combines assets that the intended swap never reaches, including other pairs, inactive ranges and the wrong side of a book. Reading that headline as usable depth produces a cheap-looking venue with a weak full-size quote.

Avoid the mistake by increasing the quote from a small probe to the intended amount without changing the mints. The output rate should deteriorate gradually. A sharp bend indicates that the order crossed a thin range, emptied a bin or consumed several order-book levels.

Mint identity and trade size come before venue choice

Venue selection starts with an exact input mint, output mint and amount because symbols alone do not identify the liquidity path being compared.

Mint addresses fix the asset

SPL Token and Token-2022 mints define different on-chain assets even when their display symbols look alike. Native SOL uses 9 decimal places, so 1 SOL equals 1 000 000 000 lamports. Wrapped SOL follows the SPL token model while preserving 9 decimals. Solana USDC uses 6 decimal places, making 1 USDC equal 1 000 000 base units. Quote systems take those integer units, and one misplaced decimal turns a depth test into a completely different trade size.

Full-size quotes expose the curve

Run back-to-back quotes at 25%, 50% and 100% of the intended input. Compare the effective output per input unit, not merely the displayed exchange rate. The 100% quote owns the decision; smaller probes reveal the curve shape. Keep exact-input and exact-output modes separate because they solve opposite constraints and traverse liquidity from different starting instructions.

Raydium establishes the depth baseline by pool type

Raydium supplies the baseline by separating constant-product pools from concentrated-liquidity pools, whose headline balances translate into executable depth in different ways.

Crucially, Raydium CLMM publishes four common configurations: 0.01% with tick spacing 1, 0.05% with 10, 0.25% with 60 and 1% with 120. A smaller tick spacing permits finer liquidity placement, while the fee is removed from input before the curve prices the swap. Raydium CPMM uses 0.01%, 0.25% and 1% configurations, while Standard AMM v4 uses 0.25%.

Compare alternatives against the specific Raydium pool that holds the pair, not an average across Raydium. A CLMM with modest total liquidity can outquote a larger CPMM when capital is tightly concentrated at the active ticks. Once the trade crosses those ticks, the advantage disappears and output deteriorates faster.

Orca Whirlpools reward depth near the active tick

Orca Whirlpools deliver strong execution when liquidity providers cluster capital around the active tick, although distant positions contribute nothing to the immediate quote.

Range placement drives output

Each Whirlpool tick represents a 1-basis-point price step, equal to 0.01%, across tick indices from −443636 to 443636. Liquidity changes when a swap crosses initialized ticks, so two pools with equal balances still expose different usable depth. Each Orca tick array holds 88 physical ticks in a 10 KB account, and one swap considers at most three arrays. A large price move can therefore meet an execution boundary even when more liquidity exists farther along the theoretical range.

Adaptive fees alter the net comparison

Orca fixed fees are stored in hundredths of a basis point, so a 0.01% fee rate is encoded as 100. Adaptive Whirlpools add a variable component to the base fee and cap the total at 10%. Better active depth still wins only when its extra output exceeds the fee difference, which is why net output outranks a liquidity screenshot.

Meteora DLMM turns depth into a ladder of bins

Meteora DLMM turns nearby liquidity into discrete price bins, making the inventory in the next few bins more important than the pool’s total balance.

A swap consumes tokens at the active bin price, then advances through adjacent bins as inventory runs out. The bin step fixes spacing: a step of 10 represents a 0.10% price increment because 1 basis point equals 0.01%. Dense inventory near the active bin gives a large order a shallow staircase; empty bins force the quote across wider levels. Meteora adds a base fee and a volatility-linked variable fee. DLMM deserves preference when the intended size remains inside well-funded neighboring bins after both components have entered the output calculation.

Jupiter combines depth that no single pool displays

Jupiter exposes combined depth by comparing direct, multi-hop and split routes rather than asking one pool to absorb the entire order.

Jupiter is an aggregator, so it is a different kind of Raydium alternative from Orca or Meteora. Its Metis engine routes on-chain across Solana DEXs, while JupiterZ obtains request-for-quote liquidity from market makers. A route can split 100% of the input between Raydium, Orca Whirlpools and Meteora DLMM, or pass through an intermediate asset such as SOL or USDC when the direct pair is shallow.

Composite depth helps only after every hop fee and transfer effect enters the calculation. A two-hop path touches two pricing mechanisms, and each adds its own price impact. Compare final output and route structure together. If a split route materially improves output at the same input size, no single venue held enough depth at the relevant price.

Aggregation therefore answers an execution question, not a protocol-loyalty question.

OpenBook V2 measures depth as queued limit orders

OpenBook V2 represents usable depth as queued bids and asks, so a market order moves price by consuming successive order-book levels.

The best ask supplies the first units of a buy. If that level holds less than the order, execution continues at the next ask and the volume-weighted average price worsens. Base and quote lot sizes set each market’s minimum order and price increments, while maker and taker fees remain market parameters. Compare cumulative size through the acceptable price range with AMM output. OpenBook V2 wins when queued inventory covers the trade without walking too far; total market deposits remain irrelevant.

Price impact, slippage tolerance and network cost stay separate

Price impact measures curve movement, slippage tolerance limits acceptable quote deterioration and Solana fees pay for execution; combining them obscures the real cost driver.

One basis point equals 0.01%, and 100 basis points equal 1%. Price impact compares the trade’s average execution with a reference price before the trade. Slippage tolerance instead creates a minimum output. Setting it to 50 basis points permits up to 0.50% deterioration from the quote during execution, but it does not improve pool depth or the quoted rate.

Solana charges a base fee of 5 000 lamports per signature, while the maximum requested compute-unit limit is 1 400 000 per transaction. The optional priority fee multiplies the requested compute-unit limit by the compute-unit price, then divides by 1 000 000 to convert micro-lamports into lamports. A direct pool and a complex route may have nearly identical price impact yet different network costs, so compare those lines separately.

The winning route is the one with the strongest net output under an acceptable execution limit.

Constant-product math reveals the reserve penalty

Constant-product math shows exactly why a trade becomes expensive relative to reserves, isolating the curve effect before routing and network costs enter, as set out in Raydium walkthrough.

The reserve ratio sets the starting quote

For reserves x and y, a CPMM preserves x × y = k. The spot ratio starts at y/x. After input Δx enters, the remaining y reserve becomes k/(x + Δx), before accounting for the swap fee. Larger Δx relative to x pushes the average execution farther from the starting ratio. Doubling both reserves while holding the trade fixed reduces that displacement.

One worked example isolates impact

A hypothetical, unit-free CPMM begins with x = 10 000 and y = 1 000 000, so k = 10 000 000 000 and the starting ratio is 100 y per x. Adding 100 x leaves 990 099.0099 y in the pool and returns 9 900.9901 y before fees. A linear quote at the starting ratio would return 10 000 y. The curve therefore removes 99.0099 y from that linear expectation, producing about 0.9901% price impact. The calculation deliberately excludes fees, isolating reserve depth as the single changing mechanism. The same method applies to a Raydium CPMM, while concentrated pools require tick-by-tick liquidity instead of one reserve equation.

Splitting the same 100-unit input into immediate sequential swaps against the unchanged pool reaches the same final reserve state, then pays repeated transaction costs. Splitting across independent pools changes the available reserves and can improve output. That distinction makes route diversity valuable.

A depth-first checklist identifies the stronger route

A depth-first checklist selects among Raydium alternatives by matching one intended trade to executable liquidity, net output and route constraints.

Use the quote itself as the decision record. Pool type explains the shape, but the final comparison must preserve identical mints, direction, amount and tolerance across candidates.

• Confirm the exact input and output mint addresses before comparing any symbols.
• Quote 100% of the intended amount, then repeat at 50% to expose nonlinear depth loss.
• Record whether execution is direct, split across venues or routed through an intermediate token.
• Compare final output after pool fees separately from minimum received and Solana transaction cost.
• Prefer the route whose liquidity remains strong across the full size, not only the first price level.

Orca Whirlpools suits trades covered by dense active ranges. Meteora DLMM fits sizes supported by nearby bins. Jupiter finds composite depth, while OpenBook V2 exposes queued liquidity and limit-price control. Raydium remains the baseline whenever its relevant pool offers the strongest net quote. The best alternative is therefore determined per mint pair, direction and order size.