The headline deliverable was Musashi Dojo, a public testnet for Ouroboros Leios that went live on June 23. It marks the first time the Leios protocol has run on a live, publicly accessible network rather than a controlled local simulation. Leios is designed to layer on top of Cardano's existing Ouroboros Praos consensus engine, adding larger "endorser blocks" during high-demand periods to push throughput significantly higher. IO's stated target is a 5x to 20x increase in base-layer capacity. Independent projections are more aggressive, modelling throughput rising from roughly 4.5 KB per second today to as much as 200 KB per second, with transactions per second potentially climbing from around 10 to more than 1,000. Those figures are projections, not guarantees, and the mainnet hard fork IO is targeting for as early as November 2026 depends on governance approvals and Musashi Dojo completing all five of its test phases.

"Musashi Dojo is live! The Leios public testnet is here. Huge credit to IO's Leios team, and to the collaborators who made this possible," IO wrote on June 23. The project involved Cardano Foundation, Intersect, Sundae Labs, Tweag, and Blink Labs. The five-phase programme, named after the chapters of Miyamoto Musashi's Book of Five Rings (Earth, Water, Fire, Wind, and Void), runs from basic protocol validation through adversarial stress testing before advancing to mainnet preparation.

ADA's price told a different story on launch day. The token fell 5.15% in the 24 hours following the Musashi Dojo announcement, dipping below $0.152 and trading near a market cap below $6 billion. By mid-July it had recovered to the $0.16 to $0.165 range. More than 60% of circulating supply remains staked, and wallets holding at least one million ADA collectively control 67.5% of supply, according to on-chain data aggregators (specific sources were not identified in the underlying research). Cardano's DeFi total value locked sits at roughly $94 million to $126 million depending on the aggregator consulted, a range that reflects methodological differences in how each platform measures liquidity. DefiLlama, the canonical reference for cross-chain comparisons, places Cardano around 27th among tracked blockchains at time of publication. The gap between engineering output and token price is a recurring tension in Cardano coverage this year.

On the node side, Cardano Node 10.7.1 and 11.0.1 were both released in Q2. Node 11.0.1 introduces support for the PV11 intra-era hard fork, codenamed Van Rossem, which requires sign-off from decentralised representatives (DReps), stake pool operators (SPOs), and the Constitutional Committee before it can take effect. The same release completed beta support for the LSM (Log-Structured Merge) storage backend. Operators using the OnDisk configuration now need only 8 GB of RAM rather than 24 GB, a reduction of roughly 65%. That number carries real weight for node operators in East and West Africa, where affordable server hardware commonly tops out at 8 to 16 GB. Previously, the 24 GB requirement was pricing out many community operators in those regions. Also completed in Q2: beta support for the KES agent and the full retirement of the legacy tracing system.

Elsewhere across the stack, Hydra shipped two node releases in Q2, including v2.2.0, which introduced Partial Fanout. Partial Fanout removes a longstanding ceiling on how many unspent transaction outputs (UTxOs) a Hydra head (Cardano's off-chain scaling layer) can contain. Previously, the head's maximum state was constrained by what could fit into a single on-chain closing transaction. Partial Fanout allows closure across multiple transactions, each verified by a BLS accumulator membership proof, opening Hydra to use cases like decentralised exchanges and games with large amounts of on-chain state. On the smart contract side, Plutus added a native Value type under CIP-153, letting on-chain scripts handle multi-asset values directly rather than through workarounds. For DeFi developers building token-heavy protocols, this reduces transaction costs and script complexity. Mithril's zero-knowledge proof work moved from prototype to testnet, with the team deploying a recursive SNARK inside the aggregate signature scheme. The Cardano High Assurance programme delivered a formalised ledger specification, the Universal Annotation Language spec, a property-based testing library, and a static-analysis rulebook.

Two programmes did not survive the quarter. IO wound down both Acropolis, a Rust-based alternative node implementation, and a Tiered Pricing mechanism in April, returning ₳1.4 million and ₳2.7 million respectively to the treasury. "We'd rather return funds and be straight about why than quietly continue work that no longer makes sense," IO stated. Acropolis delivered real infrastructure before its cancellation: the project produced a Data Node that reduced blockchain sync time from several days to roughly one hour. The wind-down reflects a strategic pivot toward chain abstraction, which IO described as hiding blockchain complexity so developers can build apps that feel like regular web software, rather than a failure of delivery. On Tiered Pricing, the logic was blunter: Leios handles fee prioritisation differently, making the old design obsolete before it was finished. All Acropolis code was open-sourced for community use. The quarter also saw DReps ratify IO's Cardano Vision 2026 research agenda with 74.96% approval. The Voltaire treasury, which now holds more than $1 billion in ADA, funds allocation through that same governance process, meaning African and South Asian delegators who participate in DRep voting have a direct say in what gets built next.

Cardano's South Asian ecosystem saw notable activity in the same period. The Cardano India Developers Community Hub continued expanding its network of regional contributors, and EMURGO's hackathon reached its Bangalore finale, drawing developers working across fintech and supply-chain verticals. The Cardano Summit 2026 is scheduled for Singapore on October 5 and 6, making it a focal point for the wider Asia-Pacific community. The remittance corridor adds further urgency to the regional picture: India receives approximately $120 billion per year in inbound remittances, a use case where Cardano's low-fee settlement layer has a credible on-ramp. Stablecoins such as USDCx are also attracting attention in high-inflation markets including Nigeria, Ethiopia, and Zimbabwe, where dollar-denominated settlement offers meaningful purchasing-power protection for users already active in the Cardano ecosystem.

The immediate test for Q3 is whether Musashi Dojo can advance through its remaining phases on schedule. A successful Leios mainnet hard fork as early as November would represent a significant milestone for the network's base layer. Whether the token market notices before then is a separate question.