Cardano Developer Seeks Reduced Funding of $46.8 Million for Expansion and DeFi Integration

Input Output, the company responsible for building and maintaining the Cardano blockchain, is seeking a substantially reduced amount of funding from the project's community treasury. The company has submitted nine proposals totaling $46.8 million for 2026, marking a notable decrease from the $97.5 million requested in 2025. Several of these proposals focus on enhancing Cardano's scalability to increase its transaction processing capacity, as well as expanding into the realm of Bitcoin DeFi. Cardano, similar to other major blockchains, operates a shared pool of funds generated by network fees, which are allocated towards development work through a voting process by community representatives. Historically, Input Output has been the largest recipient of these funds due to its significant role in employing engineers who build the underlying software. However, the reduced funding request signifies the first concrete step in a plan to gradually decrease the company's dependency on community funds. Input Output now aims to reduce its annual funding request each year, with the ultimate goal of becoming self-sufficient through its own revenue, allowing community funds to support a broader range of smaller engineering groups. By the end of 2026, the company expects smaller, specialized teams to take over the majority of its in-house work, including firms such as VacuumLabs and Midgard Labs that focus on specific layers of the Cardano software. The nine proposals can be grouped into two primary themes. The first theme involves a consensus upgrade called Leios, which Input Output claims will significantly increase Cardano's transaction processing capacity, potentially by 10 to 65 times, with a target of over 1,000 transactions per second. This would position Cardano as a competitive chain, comparable to Solana and the fastest Ethereum layer-2 networks in terms of throughput. Leios is scheduled for a test release in June, with full deployment expected by the end of the year. The second flagship proposal involves a system called Pogun, designed to introduce Bitcoin-based decentralized finance to Cardano. In practice, this would enable bitcoin holders to borrow and earn yield on their holdings through Cardano without needing to relinquish custody to a centralized intermediary. Pogun's lending component is targeted for public release in the second quarter. Smaller proposals cover various improvements, including enhancements to Cardano's smart contract engine, security testing infrastructure, developer tools, and expanded API services. Each proposal outlines specific delivery leads and ties funding to delivery milestones, rather than releasing funds upfront. This approach is similar to paying a contractor in stages as different parts of a project are completed, rather than providing the full budget at the outset. Voting on these proposals is set to open and will run through May 24, with decisions made by approximately 1,000 elected delegates known as DReps, who represent ADA holders. The outcome of this vote will test whether Cardano's governance treats Input Output like any other grant applicant or continues to approve its requests based on deference. Last year's $97.5 million proposal was approved, but since then, the Cardano Foundation has taken over the project's grant-funding arm, and Intersect, the governance organization, has assumed stewardship of core Cardano software. These shifts mean that alternatives to Input Output now exist, which could influence the voting outcome. Input Output also highlighted progress within the ecosystem, including the launch of a new Cardano stablecoin, USDCx, which reached 14.6 million tokens in circulation within weeks of its launch. Total assets deposited on Cardano, a key measure of network usage, increased from $137.5 million to $142.7 million over the same period. The outcome of the vote will signal how much the Cardano community's perspective has shifted, now that there are tools in place to fund development without relying on Input Output.