The last piece of the Starknet scaling buffet is finally here: L3s, battle-tested and production-ready.

The latest real-world test by @karnotxyz proved that the L3 building block is now fully ready; both in terms of performance and UX.

With a simple mining game (gmine) built as a L3, they achieved:

> 400 sustained TPS over 14 hours (smashing the L2Beat record with 220 sustained UOPS over a full day)

> Full L2<>L3 UX abstraction: no bridges, no network switching, no new wallets. Players didn’t even realize they were using a new chain.

Builders can now create purpose-built environments with maximum scalability and flexibility, without suffering the usual UX nightmares of fragmentation.

With this, the Starknet scaling buffet is now complete, offering every option developers need, combining ultra-performance with a web2-like UX. They can build and deploy:

> Directly on Starknet L2 for maximum composability

> As a Starknet L2 appchain for maximum flexibility

> As a Starknet L3 for hyper-scalability and maximum flexibility

Scaling without the pain is here, and it’s happening on Starknet.