The Bank for International Settlements’ Project Agorá delivered a prototype around May 27, 2026, showing that tokenized wholesale cross-border payments can settle in seconds. The experiment brought together seven central banks and more than 40 regulated financial institutions, testing a unified programmable ledger for tokenized central bank money and commercial bank deposits.
The result matters for network architects and node operators because it replaces sequential correspondent flows with coordinated atomic settlement. Compliance checks are embedded directly into transaction execution, reducing settlement risk and lowering the need for pre-funded liquidity across borders.
Unified Ledgers Change Settlement Architecture
The prototype centered on three architectural shifts. A shared programmable ledger combined messaging, reconciliation and final settlement, collapsing functions that normally sit across separate systems into one execution environment.
Project Agorá also tested the coexistence of tokenized central bank reserves and tokenized commercial bank liabilities on the same platform. That structure enabled atomic multi-currency settlement, meaning both legs of a transaction complete together or the payment cancels entirely.
Smart contracts handled compliance and screening logic in parallel with payment execution. That design compresses sequential workflow checks into a single transaction flow, helping explain how the prototype achieved near-instant settlement performance.
Node Operations Face Higher Performance Demands
For node operators and DevOps teams, the model changes infrastructure assumptions. A shared programmable ledger with embedded smart contracts increases execution-layer load, requiring predictable block propagation and low inter-node latency.
Client diversity and resilience will also be critical. Multiple implementations are needed to reduce single-client failure risk, while preserving Byzantine fault tolerance during peak transaction demand.
Continuous 24/7 settlement raises availability and synchronization requirements. Nodes will need sustained bandwidth, faster state synchronization and real-time monitoring, especially when stalled or conflicting transactions could degrade throughput.
The shift away from pre-funded nostro balances reduces liquidity footprint but raises the importance of ledger finality. Deterministic settlement at the infrastructure layer becomes essential, because liquidity efficiency depends on confidence that execution will complete reliably.
After prototype delivery, Project Agorá moved toward a real-value testing phase expected to run for roughly six months, with the Bank of Canada and additional financial participants joining. The project remains experimental and has no production deployment timeline, but its findings will inform wholesale CBDC and tokenized-payment pilots.
For implementers, the next tasks are practical: define latency targets, client interoperability tests and bandwidth profiles. Agorá’s prototype suggests tokenized wholesale payments can reduce settlement latency, while shifting operational pressure toward execution performance, resilience and continuous availability.
