# Cardano Merges Leios Core Code, but 15x Remains a Target

Published 2026-10-26 · Crypto.world Team · Crypto.world

> Cardano developers merged Praos2 and Leios networking components into the consensus codebase, while work on selecting and fetching endorser blocks continues. Input Output’s 10–15x capacity goal depends on testing, pool participation and later governance votes.

Cardano’s 15-fold capacity figure is a target, not a result of the Leios code now entering its consensus branch. The first planned step is roughly three to five times today’s Praos ceiling; Input Output says later parameter updates could take capacity to 10–15 times that level, subject to network testing and governance approval.

Source [github.com](https://github.com/IntersectMBO/ouroboros-consensus/pull/2386)

The code milestone came in stages. Pull request #2386, which introduced Praos2—Praos with the Leios overlay—merged into the IntersectMBO consensus repository on October 9. In its October 23 development report, the consensus team said LeiosNotify clients and LeiosFetch servers had also moved into the main branch. Work on selecting and fetching endorser blocks remains ahead.

## Three components are in; block selection is not

Praos2 provides the consensus code for combining Cardano’s existing Ouroboros Praos protocol with the Leios extension. LeiosNotify and LeiosFetch handle parts of the network communication: nodes can learn about Leios data and retrieve it. The development report says the prototype has also gained stricter checks on peer announcements and certificates.

Endorser blocks are central to the upgrade. Rather than placing all additional transaction data in an ordinary Praos block, Leios uses endorser blocks that carry transaction references; participating pools validate and certify them. But the development team still needs to implement the selection and retrieval work that determines which endorser blocks nodes fetch. The merged components therefore mark progress in implementation, not a completed, end-to-end mainnet upgrade.

## Why 15x requires more than a code merge

Input Output puts Praos’ current capacity ceiling at about 4.5 kilobytes of transaction data per second. Its initial Leios target is three to five times that figure. At least two later protocol-parameter updates aim for roughly 10–15 times today’s capacity, with each increase informed by results from the preceding stage.

“The hard fork is the beginning of the rollout, not the point when Cardano suddenly runs at full Leios capacity,” Ivan Irakoze of Input Output writes in the company’s October 20 rollout post. The sequence it describes includes a constitutional update, the Dijkstra hard fork, registration of pool voting keys, mainnet load exercises and subsequent parameter changes. The later capacity increases would not require another hard fork, but they would need governance approval.

Capacity is not the same as activity. Input Output says the higher limits depend on actual network demand, and its post cites sustained traffic of about 50–70 TxkB/s as the level CIP-164 estimates would be needed for fee income to reach break-even. More room for transactions does not, by itself, bring those transactions to Cardano.

## Dijkstra’s testnet has not switched eras

Intersect’s October 24 update says the public Dijkstra testnet, respun on October 21, was still running Conway protocol version 11 on node version 11.1.3. A move to Dijkstra protocol version 12 awaits a node 11.2 prerelease and a hard-fork initiation action. The testnet is available to participants, but it is not yet testing Dijkstra’s new protocol rules.

Governance and pool readiness add further gates. Intersect says the CAP-12 constitutional amendment reached a fourth draft, with items such as parameter values and pledge-leverage limits still open. Input Output estimates that BLS keys must cover roughly 90–95% of delegated stake before Leios can activate; certificates require votes representing 75% of active stake. Pools without those keys can continue producing ordinary Praos blocks, but cannot vote on endorser blocks.

For now, the confirmed gains are code in the repository and a broader public testnet preparation process—not a 15-fold increase on Cardano mainnet. Endorser-block selection and retrieval, the Dijkstra protocol transition, pool-key registration and the votes for later parameter changes remain outstanding.

## Sources

1. Original source [IntersectMBO GitHub repository](https://github.com/IntersectMBO/ouroboros-consensus/pull/2386) github.com

Other pages read

2. [Essential Cardano](https://www.essentialcardano.io/development-update/weekly-development-report-2026-10-23) essentialcardano.io
3. [Input Output](https://www.iog.io/news/leios-is-coming-how-it-reaches-cardano-mainnet) iog.io
4. [Intersect](https://www.intersectmbo.org/news/intersect-weekly-update-134-oct-24-2026) intersectmbo.org

## Related stories

- [Cardano Copper Reserve Draws 0.24% DRep Support in Early Tally](https://crypto.world/news/cardano-copper-reserve-02-percent-drep-support)
- [Intersect Proposes Seven Domains for Cardano’s 2027 Budget](https://crypto.world/news/intersect-cardano-2027-budget-seven-domains)
- [Cardano Finds 32x Storage-Latency Gap in Leios Tests](https://crypto.world/news/cardano-leios-storage-latency-dijkstra-benchmarks)
- [Intersect Tells Cardano Pool Operators to Patch Linux Flaw](https://crypto.world/news/intersect-cardano-pool-operators-linux-cve-2026-98432)
- [Cardano Assigns CIP-0206 to Quantum-Resistant Plutus Signatures](https://crypto.world/news/cardano-cip-0206-quantum-resistant-plutus-signatures)

## Explore

[More articles](https://crypto.world/news)

---

Canonical: https://crypto.world/news/cardano-leios-code-merged-15x-target
