Cardano founder Charles Hoskinson has made the network's Leios throughput upgrade the final requirement in his personal definition of a complete Cardano. In an October 17 exchange on X, he said the network he founded would be complete only after the upgrade reaches mainnet. The comment defines his own completion standard rather than serving as a formal release declaration, and it arrived while the upgrade is still being tested on a public network.

Leios remains on a public testnet, where its first phase peaked at 26.8 TxkB/s — a figure that measures transaction data throughput rather than transactions per second. In the same exchange, Hoskinson also teased an undeveloped concept for Midnight, the privacy-focused partner network, calling it the magnum opus of his career while supplying no technical design, launch date, or product specification.

Leios Framed as the Final Piece

Hoskinson's reply did not change Cardano's roadmap or create a new governance action. It instead positioned Leios as the last major element in his original design. In setting out that standard, he cited decentralized consensus, on-chain governance, continuous development, eUTXO, Plutus, scalability, formal methods, and independent node clients as components of a finished network, alongside a system able to keep improving through its own governance processes.

The statement therefore concerns completeness in the founder's view. It does not confirm that every future feature, application, or engineering task has ended. Hoskinson noted that millions of people already hold ADA, which makes the final milestone relevant beyond the engineering team. He paired the remarks with a preview of a separate project:

I'm working on a concept that only midnight can bring to the cryptocurrency space that solves most of the problems. Almost have all of it designed conceptually. It's the magnum opus of my career

— Charles Hoskinson (@IOHK_Charles) October 17, 2026

Midnight enters the picture separately from Leios. Hoskinson said he had almost finished designing, at a conceptual level, something only Midnight could bring to the cryptocurrency space, and that it could address many of the industry's problems. The post contained no architecture, feature list, timeline, or governance proposal. That leaves the announcement a personal preview rather than a confirmed product release. Neither the Midnight tease nor the Leios milestone establishes an immediate effect on ADA.

How the Leios Design Works

Cardano Leios is an extension of Ouroboros designed to expand throughput without replacing the Praos chain. Under the design, block producers publish larger endorser blocks beside regular Praos blocks. A stake-weighted committee verifies these blocks and creates a certificate for their contents, leaving the underlying proof-of-stake chain in place.

The committee's weight is determined by staking share, and certified endorser transactions enter the ledger through the existing Praos sequence. The objective is to process more transaction data while preserving the consensus system already used on mainnet, rather than replacing it.

Testnet Result Needs Context Before Mainnet

Musashi Dojo launched on June 23 as the public Cardano Leios testnet. Its Earth phase, the upgrade's first public testing stage, reached a peak of 26.8 TxkB/s, compared with 4.51 TxkB/s for mainnet using Praos alone — roughly six times the current mainnet ceiling. That comparison frames the scale of the capacity gap Leios is designed to close.

TxkB/s measures transaction kilobytes processed each second. It is not a TPS reading and cannot be treated as a direct count of mainnet transactions — a distinction that matters when weighing the figure against performance quoted in transactions per second.

The trial tested the full protocol across real distances and latencies, and it stood as the first public network run rather than a laboratory simulation. The initial phase exercised the protocol in the open, covering Praos blocks, endorser blocks, stake votes, and certificates. According to the development report, the phase produced more than 127,000 blocks, 30,000 endorser blocks, and nearly 8,000 certificates.

Engineers identified five incidents during the testing period, including forks and synchronization failures. The report described them as implementation bugs rather than failures of the underlying design. Even so, the results describe the testnet environment and cannot be read as production conditions.

Road to Mainnet Runs Through Governance

Linear Leios is slated for the first phase of the Dijkstra hard fork, targeted for late 2026. The date remains subject to Cardano's on-chain governance process, under which DReps, stake pool operators, and the Constitutional Committee must ratify a hard fork action. Stake pools must then register BLS keys for Leios voting, and the rollout also depends on pool readiness.

Once the hard fork activates, parameter changes will raise throughput in steps rather than in a single network-wide jump, allowing capacity to increase gradually as the parameters set through governance are adjusted. For readers tracking progress from here, the visible checkpoints are on-chain ratification of the Dijkstra hard-fork action, the pace of BLS key registration among stake pools, and the sequence of parameter adjustments that follow activation — each a step that must clear before any change reaches mainnet throughput.