Tapu13

@Walrus 🦭/acc is an innovative decentralized storage network for blockchain apps and autonomous agents. The Walrus storage system is being released today as a developer preview for Sui builders in order to gather feedback. We expect a broad rollout to other web3 communities very soon!
Leveraging innovations in erasure coding, Walrus enables fast and robust encoding of unstructured data blobs into smaller slivers distributed and stored over a network of storage nodes. A subset of slivers can be used to rapidly reconstruct the original blob, even when up to two-thirds of the slivers are missing. This is possible while keeping the replication factor down to a minimal 4x-5x, similar to existing cloud-based services, but with the additional benefits of decentralization and resilience to more widespread faults.
The Replication Challenge
Sui is the most advanced blockchain system in relation to storage on validators, with innovations such as a storage fund that future-proofs the cost of storing data on-chain. Nevertheless, Sui still requires complete data replication among all validators, resulting in a replication factor of 100x or more in today’s Sui Mainnet. While this is necessary for replicated computing and smart contracts acting on the state of the blockchain, it is inefficient for simply storing unstructured data blobs, such as music, video, blockchain history, etc.
Introducing Walrus: Efficient and Robust Decentralized Storage
To tackle the challenge of high replication costs, Mysten Labs has developed Walrus, a decentralized storage network offering exceptional data availability and robustness with a minimal replication factor of 4x-5x. Walrus provides two key benefits:
Cost-Effective Blob Storage: Walrus allows for the uploading of gigabytes of data at a time with minimal cost, making it an ideal solution for storing large volumes of data. Walrus can do this because the data blob is transmitted only once over the network, and storage nodes only spend a fraction of resources compared to the blob size. As a result, the more storage nodes the system has, the fewer resources each storage node uses per blob.
High Availability and Robustness: Data stored on Walrus enjoys enhanced reliability and availability under fault conditions. Data recovery is still possible even if two-thirds of the storage nodes crash or come under adversarial control. Further, availability may be certified efficiently without downloading the full blob.
Decentralized storage can take multiple forms in modern ecosystems. For instance, it offers better guarantees for digital assets traded as NFTs. Unlike current designs that store data off-chain, decentralized storage ensures users own the actual resource, not just metadata, mitigating risks of data being taken down or misrepresented.
Additionally, decentralized storage is not only useful for storing data such as pictures or files with high availability; it can also double as a low-cost data availability layer for rollups. Here, sequencers can upload transactions on Walrus, and the rollup executor only needs to temporarily reconstruct them for execution.
We also believe Walrus will accompany existing disaster recovery strategies for millions of enterprise companies. Not only is Walrus low-cost, it also provides unmatched layers of data availability, integrity, transparency, and resilience that centralized solutions by design cannot offer.
Walrus is powered by the Sui Network and scales horizontally to hundreds or thousands of networked decentralized storage nodes. This should enable Walrus to offer Exabytes of storage at costs competitive with current centralized offerings, given the higher assurance and decentralization.
The Future of Walrus
By releasing this developer preview we hope to share some of the design decisions with the decentralized app developer community and gather feedback on the approach and the APIs for storing, retrieving, and certifying blobs.
In this developer preview, all storage nodes are operated by Mysten Labs to help us understand use cases, fix bugs, and improve the performance of the software.
Future updates to Walrus will allow for dynamically changing the set of decentralized storage nodes, as well as changing the mapping of what slivers are managed by each storage node. The available operations and tools will also be expanded to cover more storage-related use cases. Many of these functions will be designed with the feedback we gather in mind.
Stay tuned for more updates on how Walrus will revolutionize data storage in the web3 ecosystem.
What can developers build?
As part of this developer preview, we provide a binary client (currently macOS, ubuntu) that can be operated from the command line interface, a JSON API, and an HTTP API. We also offer the community an aggregator and publisher service and a Devnet deployment of 10 storage nodes operated by Mysten Labs.
We hope developers will experiment with building applications that leverage the Walrus Decentralized Store in a variety of ways. As examples, we hope to see the community build:
Storage of media for NFT or dapps: Walrus can directly store and serve media such as images, sounds, sprites, videos, other game assets, etc. This is publicly available media that can be accessed using HTTP requests at caches to create multimedia dapps.
AI-related use cases: Walrus can store clean data sets of training data, datasets with a known and verified provenance, model weights, and proofs of correct training for AI models. Or it may be used to store and ensure the availability and authenticity of an AI model output.
Storage of long term archival of blockchain history:Walrus can be used as a lower-cost decentralized store to store blockchain history. For Sui, this can include sequences of checkpoints with all associated transaction and effects content, as well as historic snapshots of the blockchain state, code, or binaries.
Support availability for L2s: Walrus enables parties to certify the availability of blobs, as required by L2s that need data to be stored and attested as available to all. This may also include the availability of extra audit data such as validity proofs, zero-knowledge proofs of correct execution, or large fraud proofs.
Support a full decentralized web experience: Walrus can host full decentralized web experiences including all resources (such as js, css, html, and media). These can provide content but also host the UX of dapps, enabling fully decentralized front- and back-ends on chain. It brings the full "web" back into "web3".
Support subscription models for media: Creators can store encrypted media on Walrus and only provide access via decryption keys to parties that have paid a subscription fee or have paid for content.
#walrus $WAL

@Plasma There is a quiet kind of engineering ambition inside the idea of a blockchain built only for money that does not move with the same drama as DeFi or NFTs. Plasma is one of those efforts: it does not try to be everything to everyone. It chooses a single mission and treats it like a product requirement rather than a slogan. From the perspective of someone who moves cash for a living, that focus changes how you judge success. Speed is important, but reliability, predictable cost, censor resistance and an obvious path to integrate with existing financial plumbing matter more. Plasma is interesting because it designs for those operational realities first.
Plasma’s technical commitments read like a checklist of what a payments engineer would ask for at three in the morning. Full EVM compatibility lowers friction for developers who already know Ethereum tools, meaning wallets and contracts do not need reinvention. A consensus layer tuned for near-instant finality reduces the practical headache of reconciliation. And the insistence on enabling zero-fee transfers for widely used stablecoins reframes a UX problem as a protocol feature: users should not have to hold a speculative token simply to move a dollar. Taken together, these choices are not just convenience; they are a statement about what money-on-chain should feel like when it is used day after day.
There is another subtle but important design decision that changes the political and operational calculus: anchoring security to Bitcoin. To many institutions, security is not merely a math proof. It is a narrative of neutrality. When a settlement layer ties itself to Bitcoin’s security assumptions, it is signaling something practical. It accepts a conservative story about where trust should sit, and that can make partnerships with payment providers or regulated institutions easier to imagine. That does not magically solve all regulatory headaches, nor does it buy trust for projects that are opaque about reserves or governance. It does, however, offer a different architecture for resisting unilateral censorship and for building messaging that compliance teams can parse.
None of this eliminates the real tensions. Building a payments chain around a particular set of stablecoins raises questions about concentration, issuer risk and interoperability. If a chain optimizes for one dominant token, what happens when that issuer changes policy or gets entangled in sanctions? If zero-fee transfers depend on offchain commercial arrangements or privileged privileges from issuers, what does neutrality mean in practice? Sensible project design should make these tradeoffs explicit and give clear operational paths for fallback, auditability and dispute resolution. Users and institutions do not buy technology; they buy predictable outcomes. Protocol teams that understand that will have a better chance at sustainable adoption. Documents and launch messaging alone do not achieve that. Real trust is made in operational processes, audits and consistent incident handling.
If you zoom out, the most interesting question is not whether Plasma will win as a brand. The more consequential possibility is that we are seeing the early formation of a payments layer that other rails, wallets and institutions can plug into for settlement without adopting all the social dynamics of speculative markets. That matters for markets where stablecoins are already used as money, for remittances where speed and certainty materially lower friction, and for institutional flows that prefer predictable fee models. The test will not be clever tokenomics or marketing but whether real-world payments teams can replace outdated corridors with a system that reconciles with banks, custodians and treasury systems without adding new kinds of operational risk.
So read the product roadmap as a promise to operational teams, not a marketing brochure.The promise is straightforward: make dollar transfers onchain feel like internal bank wires, with the safeguards and auditability that institutions require. Achieving that requires smart engineering, yes, but even more importantly it demands a culture of continuous operations, clear legal boundaries and a willingness to be judged by uptime and reconciliation records rather than by token price. If Plasma and projects like it can lock attention on those criteria, the most meaningful change might not be a new token that people hype but a quieter shift in how digital money moves around the world.
#Plasma $XPL

@Walrus 🦭/acc I didn’t approach Walrus with enthusiasm. That reaction has been worn down over time by a long line of decentralized storage and privacy-focused DeFi projects that sounded right and behaved poorly. The category has learned to speak convincingly about censorship resistance and data ownership, but far less convincingly about costs, reliability, and day-to-day usage. What surprised me with Walrus was not a breakthrough announcement or a dramatic claim, but a slow realization that this system felt designed with skepticism in mind. It didn’t assume goodwill from the market. It assumed friction, failure, and pressure. And in crypto, that assumption alone already sets it apart.
The design philosophy behind the Walrus Protocol reflects that realism. Walrus treats privacy and decentralization as constraints that have to survive contact with reality, not as ideals that excuse weak trade-offs. Built on the Sui Blockchain, it relies on erasure coding and blob storage to distribute large files across a decentralized network. These are not novel ideas, and that is precisely the point. Erasure coding reduces the need for full replication while still allowing data to be recovered when parts of the network fail. Blob storage accepts something many blockchains avoid acknowledging: most real-world data is large, unstructured, and not meant to live directly on-chain. Walrus doesn’t try to bend data into a cleaner shape. It builds infrastructure that works with how data actually behaves.
What makes Walrus notable is how intentionally limited its scope feels. It is not trying to become a universal storage layer or a replacement for every cloud provider. The focus is narrower and more practical: cost-efficient, censorship-resistant storage paired with private transactions that can support real applications today. WAL, the native token, exists to support governance, staking, and participation, but it is not positioned as the core product. That distinction matters. Many infrastructure projects faltered because the token became the main event while usage remained optional. Walrus appears to operate under the opposite assumption: if people don’t actually need the storage, no incentive design will manufacture sustainable demand.
There is also a noticeable absence of performance theater. No promises of infinite scalability. No exaggerated throughput numbers divorced from context. Instead, the emphasis is on predictable behavior and clear trade-offs. Lower costs compared to fully replicated decentralized storage models. Simpler integration paths for developers who don’t want to re-architect their systems. An acceptance that efficiency matters more than spectacle once infrastructure leaves the demo stage. In an industry that often mistakes complexity for progress, Walrus’s simplicity reads less like compromise and more like confidence.
From the perspective of someone who has watched several infrastructure cycles rise and fall, this approach feels familiar in the right way. I’ve seen storage networks that looked elegant collapse when real data volumes arrived. I’ve seen privacy-focused systems become unusable once fees and latency met actual users. Walrus feels shaped by those lessons. It does not pretend decentralization removes cost pressure or operational friction. It accepts that privacy, performance, and efficiency will always pull against each other, and it tries to manage that tension rather than deny it. That mindset usually comes from experience, not optimism.
The open questions remain, and they should.Can Walrus maintain its cost efficiency as adoption grows and storage demands scale? Will enterprises trust decentralized storage enough to move sensitive workloads, even with privacy guarantees in place? How will governance evolve as WAL holders balance personal incentives with long-term network sustainability? These questions echo the same challenges that have followed decentralized infrastructure for years, from scalability limits to the unresolved trade-offs of the trilemma. What makes Walrus interesting is not that it claims to have solved these problems, but that it seems built with the expectation that they will persist. If it succeeds, it will be because people continue to use it when there is no excitement left to carry it. And in a space crowded with promises, that kind of quiet usefulness feels like genuine progress.
#walrus $WAL

@Walrus 🦭/acc I didn’t come across Walrus expecting to be persuaded. After spending years watching decentralized storage and privacy-first DeFi projects overpromise and quietly stall, skepticism feels like the default setting. Most of these systems sound compelling until you imagine them under real pressure, with large files, real users, and budgets that actually matter. What caught me off guard with Walrus was how little it seemed interested in selling a grand vision. There was no dramatic declaration about fixing the internet or redefining trust. Instead, there was a quieter signal. This looked like a project designed by people who assume the real world will push back hard, and who decided to build accordingly. That realization didn’t erase my doubts, but it did make me curious in a way that felt earned.
The design philosophy behind the Walrus Protocol reflects that realism. Walrus treats privacy and decentralization less like ideals and more like constraints that have to survive contact with cost, latency, and failure. Built on the Sui Blockchain, the protocol uses erasure coding and blob storage to distribute large files across a decentralized network. These are not flashy choices. They are practical ones. Erasure coding exists because it allows data to be reconstructed even when parts of the system fail, without the expense of full replication. Blob storage exists because most modern data is large, unstructured, and simply does not belong directly on-chain. Walrus doesn’t try to force data into an elegant blockchain shape. It reshapes the infrastructure to fit how data already behaves.
What stands out most is how deliberately narrow the scope is. Walrus is not trying to replace every cloud provider or become the storage layer for all of Web3. Its ambition is smaller, and because of that, more believable. Cost-efficient, censorship-resistant storage. Private transactions that support real applications. WAL, the native token, supports governance, staking, and network participation, but it doesn’t pretend to be the main product. That choice matters. Many infrastructure projects failed because speculation around the token overshadowed actual usage. Walrus seems built around the assumption that if people don’t genuinely need the storage, no incentive design will create lasting demand. The system has to earn its place through use, not narrative.
This emphasis on practicality over hype feels almost contrarian in crypto. There are no exaggerated claims about infinite scalability or world-changing throughput. Instead, the value proposition is grounded. Lower costs than fully replicated decentralized storage models. Predictable behavior under load. Trade-offs that are visible rather than hidden. For developers and organizations, that clarity matters more than theoretical peaks. Simplicity, in this case, feels less like a limitation and more like confidence. It suggests a team that understands that infrastructure lives or dies on reliability, not ambition.
From the perspective of someone who has watched several infrastructure cycles come and go, this approach feels familiar in the best way. I’ve seen storage networks that worked beautifully in controlled environments collapse when real data volumes arrived. I’ve seen privacy-focused systems become unusable once fees and latency met actual demand. Walrus feels shaped by those lessons. It doesn’t pretend decentralization removes economic pressure or operational friction. It accepts that privacy, cost, and performance exist in constant tension and tries to manage that balance rather than deny it. That mindset usually comes from experience, not optimism, and experience tends to matter once the novelty fades.
The open questions remain, and they should.Can Walrus maintain cost efficiency as adoption grows and data demands become less forgiving? Will enterprises trust a decentralized storage layer enough to move sensitive workloads, even with strong privacy guarantees? How will governance evolve as WAL holders balance personal incentives with long-term network sustainability? These questions echo the broader challenges that have defined decentralized infrastructure for years, including scalability limits and the unresolved trade-offs of the trilemma. What makes Walrus interesting is not that it claims to have solved these problems, but that it seems designed with the expectation that they will persist. If it succeeds, it will be because people keep using it long after the excitement fades. And in an industry crowded with promises, that kind of quiet usefulness feels like real progress.
#walrus $WAL