Cardano boots first Ouroboros Leios public testnet under the name Musashi Dojo

Input Output Global brought the first public testnet for Ouroboros Leios online on June 23, 2026 under the name Musashi Dojo, kicking off a structured, multi-phase validation of CIP-0164 — the proposed Cardano protocol upgrade that would push Layer 1 throughput 30–65× above today's Praos. The testnet name borrows from Miyamoto Musashi's Book of Five Rings and partitions the program into five stages: Earth, Water, Fire, Wind, Void, running from basic protocol validation through adversarial stress before any mainnet vote.

What shipped

Musashi Dojo is the first end-to-end network running a working Linear Leios prototype, not a single-node simulation. CIP-0164, opened as PR #1078 on the Cardano Foundation CIPs repo on March 7, 2025 and revised through review since, defines the protocol. It is an overlay on Ouroboros Praos, not a replacement: block producers continue minting standard Praos blocks at the existing schedule. The new construct is a second class of block.

• Ranking Blocks (RBs) — the familiar Praos block, extended with optional fields that announce and certify a separate Endorser Block.
• Endorser Blocks (EBs) — larger blocks carrying references to additional transactions, validated by a stake-weighted voting committee before inclusion in the ledger.

The committee structure uses what the CIP calls stake-based truncation: pools are selected in descending order of stake until a cumulative target σc is reached. Membership is fixed once per epoch, with no per-EB sortition. The protocol also defines three timing windows on which security rests: equivocation detection (3·L_hdr), the committee voting period (L_vote), and the diffusion period (L_diff) that has to elapse before a certified EB hits the chain.

The numbers in the spec

CIP-0164 states the throughput target verbatim:

- Mainnet today: ~4,500 transaction bytes per second (TxB/s)
- Linear Leios target: 140,000–300,000 TxB/s
- Equivalent: ~100–200 transactions per second
- Multiplier: 30–65× over current Praos
- Required by 2029: ~36–50 TPS for SPO profitability as the Cardano Reserve depletes
- CIP status: Proposed
- Original PR: cardano-foundation/CIPs#1078 (2025-03-07)

The 2029 figure is the one that drives the timetable. Cardano's monetary expansion comes from the Reserve; as the Reserve drains, the gap between block-reward subsidies and SPO operating costs has to be covered by transaction fees. CIP-0164 cites roughly 36–50 TPS as the floor needed for staking pools to remain profitable on fees alone once the subsidy runs down.

Five phases

The Musashi Dojo testnet runs the prototype through five named stages, each a layer of testing the CIP's acceptance criteria call out — formal security reduction to Praos, parameter validation under simulation, real-world performance under load, then adversarial scenarios. The five chapters map to The Book of Five Rings:

1. Earth — protocol-level design validation. Does the spec do what it says under nominal conditions?
2. Water — parameter tuning. The L_hdr, L_vote, L_diff windows and committee target σc get fitted to mainnet-like stake distributions.
3. Fire — performance under realistic load. The 140,000–300,000 TxB/s window is the target to hit.
4. Wind — adversarial scenarios. Equivocation attempts, vote-period DoS, withholding strategies.
5. Void — mainnet-readiness checks. Audits, final stress runs, the go/no-go for a Voltaire hard-fork vote.

The earliest phase is now running. There is no public schedule yet for transitions between phases.

Where it sits in the governance pipeline

Cardano is in the middle of a separate hard-fork governance process: the van Rossem hard-fork action, submitted on June 16, 2026, is on the DRep ballot now and would raise the protocol version to 11. van Rossem is the intra-era step that prepares the chain for Dijkstra, the era that will host Leios. Leios itself will require a separate governance action — and a separate DRep ratification — before it can hard-fork onto mainnet.

That ratification is the open question. Leios development is funded: a treasury action approved earlier in 2026 unlocked roughly 27.7 million ADA from the Cardano treasury for the work, passing with ~84% DRep support. But IOG's flagship research budget — the 32.9M ADA Cardano Vision 2026 IO Research ask — closed on June 8 trailing 86.7% no, with Charles Hoskinson saying publicly that IOG will not resubmit. DReps have shown they will fund engineering and reject open-ended research grants in the same week. The Leios mainnet vote, when it comes, will land in that environment.

Pattern

Musashi Dojo is the third Cardano protocol-track milestone covered on this site in roughly seven weeks: the IO Research treasury rejection (June 8), the van Rossem hard-fork action (June 16) and now the Leios public testnet (June 23). Read together they describe a chain that is shipping engineering on a tighter cadence than it has in years, while simultaneously being told by its own electorate that the era of unconditional treasury support for IOG is over. Leios is the technical bet that has cleared the funding bar so far. The next test is whether the spec performs under the Earth-through-Void program in time for a Voltaire vote that has to clear a 66.67% DRep supermajority on the hard-fork action itself.

What to watch

1. The Musashi Dojo phase tracker — when Earth → Water → Fire transitions are announced, and whether the Fire-phase throughput hits the 140,000 TxB/s floor in CIP-0164.
2. Any updates to CIP-0164 — parameter changes during testnet are still in scope.
3. The first Leios-specific governance action on gov_action… — until one is submitted, the November mainnet target IOG has floated is informational, not on-chain.
4. The van Rossem ratification result on the Proposal Examiner. Earliest enactment: June 28, 2026.