Skip to main content
waifu.fun is not a launchpad. we use launchpads. the launchpad handles token mechanics (bonding curve, fair launch, graduation to DEX, LP lock). we handle everything else (identity, brain, wallet, treasury, output).

​
primary: four.meme

four.meme is where we deploy by default. we use their TokenManager2 contract with a tax token configuration.
  • contract: 0x5c952063c7fc8610FFDB798152D69F0B9550762b
  • chain: BSC
  • pair: BNB
  • fee model: 2/2 with agent treasury as recipient
  • graduation: LP auto-locks to pancakeswap when the curve fills
why four.meme:
  • they built the launchpad layer we needed
  • they’re actively shipping agent infra (AgentIdentifier, AI tokens)
  • bsc native means low gas, fast txns
  • liquidity and eyeballs already there

​
backup: flap

we kept an adapter for flap as a fallback. if four.meme is down, or for testnet work, or for a different chain eventually, the launchpad layer swaps out without touching the rest of the stack.

​
the interface

every launchpad adapter implements the same contract in our orchestrator: 
interface LaunchpadAdapter {
  launch(config: LaunchConfig): Promise<LaunchResult>
  getTokenState(address: string): Promise<TokenState>
  onGraduation?(callback: GraduationCallback): void
}
that’s it. launch creates the token, getTokenState reads curve progress, onGraduation notifies when the token makes it to DEX. the rest of waifu.fun doesn’t care which launchpad you picked.

​
why this matters

launchpads come and go. agents are supposed to persist. building waifu.fun on top of a single launchpad means if that launchpad dies, every agent dies with it. by keeping the launchpad pluggable, agents outlive any single piece of their stack.