D20DAO reports 470 Arc testnet requests completed within its 60-second deadline

Developer report | Documentation reviewed September 17, 2026.

D20DAO has published a September 17 benchmark report covering 470 paid randomness requests across four Arc Testnet scenarios. The provider reports that all were fulfilled within the protocol's 60-second acceptance window, with no expired requests or refunds in those runs.

The report separates light traffic from concentrated demand. Twenty sequential requests recorded a median completion time of two chain seconds. A burst of 200 requests recorded a median of 21 seconds and a maximum of 35 seconds. Under sustained arrivals of five requests per second, the reported median was five seconds and the maximum eight.

The variation is useful for developers designing a waiting screen or separating a player's action from its eventual outcome. A fast isolated request is not evidence that every request remains equally fast under load.

The report links scripts and result files at a fixed revision, together with request and fulfillment transactions. Its latency measure uses block timestamps; it also reports wall-clock observations separately. All scenarios used one keeper on testnet with a small callback and a 100,000-gas callback allowance.

These are provider-published measurements, not independently reproduced results, a mainnet capacity claim or a service-level guarantee. Applications still need an expiry path and a way to recover failed callback delivery.

Sources: D20DAO benchmark report and methodology, service rules.