This page has been updated for Wallets SDK V1. If you are using the previous version,
see the previous version of this page or the V1 migration guide.
The Problem with Off-Chain Key Storage
Most wallet infrastructure today stores keys off-chain, inside proprietary enclaves or distributed across MPC networks. This means:- Provider dependency: Wallets depend on a specific provider’s infrastructure to function.
- Latency overhead: Key reconstruction adds latency to every signature.
- Painful migration: Moving to a new provider means exporting keys and creating new wallet addresses.
A Different Approach
Smart contract wallets take a different approach. The wallet is a program deployed on a public blockchain — it does not live in anyone’s private infrastructure. The signer is completely separate and can be anything: a passkey on a user’s phone, an AWS KMS key in your cloud, or a recovery credential. The key stays whole. No splitting, no reconstruction, no added latency.- Security is not a promise — it is code you can read.
- Recovery is not a key-share coordination problem — it is an onchain signer rotation.
- Migration is not an export — it is a signer swap.
Why We Chose This Architecture
We chose this architecture because it is the only one that scales to the future we are building toward: wallets as ubiquitous and permanent as email addresses, but programmable, auditable, and owned by no one except the people and systems that use them.Architecture
See how the dual-layer architecture works in practice.
Signers
Explore the signer types that plug into this architecture.

