Cardano's Leios upgrade is a smaller version of the design its own research proposed

A Practical Compromise on an Ambitious Design

@Cardano's current consensus mechanism, Ouroboros Praos, deliberately leaves quiet gaps between blocks. Leios is designed as an optimistic consensus protocol that builds on Praos, with block producers simultaneously creating both a standard Praos block and a larger secondary block referencing additional transactions. Those secondary structures fill the idle bandwidth that Praos leaves unused.

The original Leios research described a more ambitious, fully concurrent design with independent block streams. Input Output's own design documentation, however, concluded that version was not directly implementable against Cardano's actual ledger. The proposal now on the table, called Linear Leios, ties each new endorser block to an ordinary ranking block instead. Secondary blocks undergo committee validation before ledger inclusion, enabling significantly higher throughput. If no endorser block is certified in a given slot, the chain continues under standard Praos rules without interruption.

Throughput Targets and the Dijkstra Roadmap

The scaled-back design gives up some of the original ambition in exchange for something that can actually be shipped. The architecture introduces additional block structures including endorser blocks, allowing more transactions to be processed and certified through stake-based committees, with Input Output targeting an initial fivefold to 20-fold increase in Cardano's base-layer throughput capacity.

The proposal targets a 10x to 65x increase in transaction capacity overall, with the official roadmap describing a move from roughly 4.5 KB/s today toward 200 KB/s, though Input Output product manager Carlos Lopez de Lara has said the mainnet rollout would start conservatively at around 2x to 5x current capacity, with the higher ceiling unlocked as demand grows.

Linear Leios sits inside Cardano's broader Dijkstra era. Cardano developers have finalised the scope of the Dijkstra era, setting out a two-phase upgrade that will first deploy Ouroboros Linear Leios and Nested Transactions before introducing the Peras consensus enhancement through a separate hard fork in 2027. Code completion is targeted for Q4 2026, although that date does not represent final mainnet activation.

As a result, Cardano can increase transaction capacity without relying on larger blocks or faster slots. The design preserves the security model that Praos already provides, keeping the protocol's existing resistance to adversarial stake intact.

Sources
CIP-0164: Ouroboros Linear Leios, Cardano Improvement Proposals
Cardano is ready to grow: Leios treasury proposal, Cardano.org
Cardano plans two-phase Dijkstra upgrade with Leios and Peras, FinanceFeeds