Pump.fun Abandons Raydium, Plans Own AMM

Generated by AI AgentCoin World
Tuesday, Feb 25, 2025 5:12 pm ET1min read

Pump.fun, a prominent Solana-based memecoin launchpad, is reportedly abandoning its partnership with Raydium, an automated market maker (AMM) on the same blockchain, and plans to launch its own AMM. This move, if confirmed, could significantly impact the crypto landscape and the revenue streams of both platforms.

The current arrangement between Pump.fun and Raydium involves the migration of tokens with sufficient liquidity to trade on the latter's platform. This has proven highly profitable for Raydium, with Pump.fun pools accounting for a substantial 36% of the DEX's $154 million in 2024 swap revenue. However, the recent appearance of an 'amm.pump.fun' URL and evidence of testing liquidity pools for its own AMM suggest a shift in strategy.

AMMs are relatively easy to set up in the open-source crypto world, and rumors of Pump.fun's aspirations to create its own have been circulating for some time. By launching its own AMM, Pump.fun could vertically integrate the token launch process and potentially boost its already impressive revenue figures. However, it's unclear how this move will impact the liquidity and value of the memecoins launched on its platform.

Pump.fun had previously been approached by Ellipsis Labs about implementing Plasma, an AMM designed to limit sandwich attacks. However, Pump.fun declined the offer, indicating a preference for developing its own AMM. The launch of an in-house AMM could provide value accrual to any future Pump.fun token, but the success of this strategy will depend on whether hardcore memecoin traders deem the new AMM liquid enough to be worth their time.

Raydium core contributor Infra has warned that Pump.fun's decision to launch its own AMM could be a strategic miscalculation, as Raydium has provided the liquidity that helped Pump tokens thrive. However, the closed-source nature of Pump.fun suggests that any future AMM launch could be a departure from its current operational model.