Storage is the starting point, proof is the business, a brief analysis of the Irys roadmap.

In "DataFi Project Overview: From Mechanisms to Implementation, Who is Building a New Order in Data Finance? (Part One)", Portal Labs first conducts research around the DataFi track and the OpenLedger project, attempting to outline the basic mechanisms of the track and early implementation directions. Considering the complexity of project paths and technical logic, Portal Labs will subsequently publish multiple articles to study and analyze representative projects one by one.

If OpenLedger's story revolves around building a DataFi business model with AI, then Irys extends another set of business logic based on storage.

In last week's article "The Three Paths of Web3 Data: Storage, Analysis, and Assetization, Which is the Ultimate Path?" published by Portal Labs, we summarized the three paths of the Web3 data track: storage, analysis, and assetization, and judged that assetization will be a key direction for future business models. Irys is precisely at the intersection of this logic: it starts with storage and is attempting to move towards assetization of data.

About Irys

The story of Irys begins with the Bundlr Network. As a scaling tool in the Arweave ecosystem, Bundlr's early positioning was quite singular, taking on the tasks of transaction packaging and high throughput on-chain, making Arweave faster and cheaper. And it indeed did well, helping Arweave achieve over 1 billion transactions. However, as the scale of data expanded, Bundlr gradually accumulated independent value and chose to rebrand itself as Irys in 2023.

Irys has proposed the vision of a "trustworthy ledger," emphasizing that data can not only be stored but also self-verify its source and reliability, attempting to move beyond the positioning of a "tool" towards an independent programmable data chain.

The financing path also reflects this transformation: in June 2024, Irys completed a new round of financing led by Lemniscap; on August 21, 2025, it completed a $10 million Series A round led by CoinFund, bringing the total financing to nearly $20 million. Both Lemniscap and CoinFund are known for their focus on infrastructure and new narratives. This also shows the market's expectation of Irys as the "DataFi data foundation."

The core of Irys's transition lies in the fact that it no longer views data merely as "archival" but instead defines it as callable and orchestratable elements. Irys draws on the developmental logic of the internet, pointing out that the early internet was stuck in the static page phase, and true prosperity comes from the application ecosystem released by programmability. Similarly, in the on-chain world, if data is merely stored in isolation, its value is limited; only when data can flow, trigger logic, and be embedded in applications does it truly become a means of production.

In the narrative of Irys, the significance of "programmable data chain" lies in making data an active element at the application layer. Compared to traditional storage protocols, it emphasizes three differences:

• Proof of credibility. Data is accompanied by verifiable sources and integrity when it is on-chain.
• Programmability. Through the virtual machine interface (IrysVM), data can directly drive contract logic.
• Long-term availability. In a multi-ledger structure, it retains permanent storage while also providing flexible terms, reducing cost burdens.

For this reason, Irys hopes to place itself in a new coordinate system—one that is different from networks like Arweave and Filecoin, which are centered around "storage", and distinct from solutions like Celestia and EigenDA, which aim for "availability"; instead, it exists as an underlying infrastructure for "integration of storage and execution".

To achieve this positioning, the core features of the Irys product design are mainly reflected in three aspects:

First, a multi-ledger architecture. Data does not directly enter permanent storage at the beginning; instead, it is first written to the Submit Ledger for verification and replication, and then enters the Publish Ledger for long-term storage. In the future, a Term Ledger will also be introduced, allowing developers and enterprises to choose storage based on time limits. This hierarchical structure not only reduces costs but also allows different types of data to find suitable homes.

Second, IrysVM. Through a set of EVM-compatible virtual machines, developers can directly read, orchestrate, and write data on-chain. This way, data is no longer a "passively existing" archive, but can directly drive contract logic, becoming input at the application layer. In other words, Irys enables stored data to have "programmability."

Thirdly, consensus and incentive mechanism. Irys adopts a combination of uPoW (useful Proof of Work) + Staking. Nodes not only need to pledge storage space (with 16TB as the basic unit), but also need to continuously submit storage integrity proofs. Nodes that fail to meet the requirements will be penalized, while qualified nodes will receive rewards. This mechanism ensures long-term availability and enhances the credibility of the data.

Around these designs, Irys provides SDK, CLI, and Gateway services for interface development, supporting multi-currency payments (ETH, USDC, SOL, etc.), and offers variable references and folder organization features, reducing the complexity of developer integration. Overall, its product logic is: starting from storage, gradually transitioning to data calling and execution, making data a "first-class citizen" on the chain.

Business Model

If the product design explains how Irys works, then the business model determines how it captures value in the market. Overall, Irys's revenue and incentive structure are still built on "storage," but on top of that, a dimension of "execution" has been added, forming a dual-layer logic.

First, storage fees.

Irys's core revenue still comes from storage. The permanent ledger adopts a one-time payment model, with pricing anchored to the physical storage costs, aiming to minimize the impact of token price fluctuations. The planned fixed-term ledger in the future will provide businesses with more flexible options, such as data that only needs to be stored for a few years instead of permanently. Through this tiered pricing mechanism, Irys aims to balance the long-term needs of Web3 applications with the cost sensitivity of enterprises.

Second, execution and invocation fees.

In IrysVM, deploying and calling contracts incurs transaction fees. This part of the logic is similar to the "Gas model" of public chains, but its uniqueness lies in the fact that execution and storage are placed within the same network. Developers no longer need to operate cross-chain when calling data; instead, they can directly complete reading, verifying, and writing on Irys. This design makes execution costs more predictable and adds an additional source of revenue for Irys beyond storage itself.

Third, miner incentives and network security.

The nodes on the supply side require both pledge storage space and must submit proof of storage integrity. Their incentives primarily come from the sharing of fees for uploading data and transaction fees. For the network, this model not only continues the basic logic of decentralized storage protocols but also directly binds economic benefits to storage availability through staking and useful work (uPoW).

Fourth, value proposition.

For enterprises, Irys's selling point is not "cheaper," but "more certain." A stable pricing model, low-friction invocation method, and multi-currency payment interface bring it closer to an enterprise-level data service logic, rather than just a technical attempt within the crypto community. The core narrative of Irys is: when data becomes a programmable element, storage itself can become the starting point of a business model, rather than the endpoint.

These abstract sources of income only become meaningful when placed in specific contexts:

• AI Field: The credibility and source proof of training data determine whether the model can be trusted. Irys provides verifiable training set sources through a "credibility ledger," and companies are willing to pay for this.
• DePIN Field: Device data on the chain requires scalable and low-cost storage. Term Ledger provides different retention periods for different device data, and automatic settlement relies on calling fees.
• NFT and Copyright: The evidence and tracing of digital works only have real value when they are "callable" on the chain. IrysVM allows royalty settlements to be directly linked to stored data, bringing execution costs.
• RWA and Compliance Auditing: When real-world assets are mapped onto the chain, the authenticity and long-term availability of data are crucial. Irys's credible ledger and long-term storage logic are precisely what enterprises are willing to pay for.

Competition and Advantage

Irys has shifted its focus from storage and is now attempting to align itself with data assetization, which also means that the competition it faces is no longer limited to the storage network itself.

In the storage field, Arweave and Filecoin are the most direct references. Arweave emphasizes extreme long-term storage with its core narrative of "one-time payment, permanent storage"; Filecoin adopts a leasing model, which is closer to the market supply and demand logic. Irys occupies a position between the two: it extends the capability of permanent storage while introducing flexible terms through a multi-ledger structure, providing more options for different types of enterprises and developers.

In the data availability layer, Celestia and EigenDA focus on "ensuring data can be accessed," widely applied in L2 and rollup scaling. Unlike this lightweight availability solution, Irys emphasizes data integrity and long-term accessibility, aiming to establish continuity between storage and retrieval.

In the data orchestration field, protocols like Ceramic and Tableland offer flexible data models and interfaces that are suitable for rapid application development, but they typically rely on external storage. Irys differentiates itself by binding storage and execution on the same chain, aiming to reduce friction caused by multi-system dependencies while ensuring long-term sustainability.

In the DataFi track, projects like OpenLedger and DataDance generally center around "data assetization" as their core logic, designing incentive models around the acquisition, analysis, and trading of personal data, allowing individual users to directly participate in the circulation and monetization of data. Irys, on the other hand, continues the B2B infrastructure path, placing more emphasis on the credibility and long-term availability of data. Compared to C-end DataFi projects, Irys serves more as a foundational support element within the DataFi ecosystem, providing storage and verification capabilities for assetization logic.

From these comparisons, it can be seen that Irys's advantages mainly focus on several aspects:

• Integration. Integrate storage and execution on the same chain to avoid the friction of cross-chain calls;
• Flexibility. By using a multi-ledger structure, it balances permanent and fixed-term storage to meet the diverse needs of different enterprises;
• Stability. A pricing mechanism anchored to the physical storage cost allows businesses to predict costs more clearly.
• Developer-oriented. SDK, Gateway, and multi-currency payment interfaces reduce the integration threshold for applications.

At the same time, Irys also faces challenges. It needs to establish differentiation in the storage market where Arweave and Filecoin have already formed scale effects, while also proving its position under the grand narrative of DataFi.

Irys's "permanent storage" feature also brings another layer of limitation: it is better suited for preserving and verifying historical data, but in scenarios that require high real-time and freshness (such as AI Agents, advertising bidding, and financial data circulation), it remains questionable whether it can provide sufficient dynamic supply. This means that even if Irys can establish an advantage in the credibility of long-term data, it still needs to find a complementary relationship with real-time data circulation solutions.

Additionally, Irys also faces uncertainties in terms of compliance. How permanent storage aligns with the "right to be forgotten" in regulations like GDPR is a question that all similar agreements cannot avoid. Although Irys has introduced more flexible storage mechanisms like Term Ledger, how to interpret the conflict between long-term preservation and user deletion rights remains a challenge that needs a clear path in a regulatory context. Meanwhile, cross-border data flow is becoming a key regulatory focus globally. If Irys is positioned as an enterprise-level infrastructure, it must confront the compliance of data storage locations and invocation paths. Furthermore, if Irys's token model is directly linked to storage and invocation, it may also touch on securities attributes or financial infrastructure standards, which poses potential pressure on its B2B expansion.

Summary

From Bundlr to Irys, this path reflects Portal Labs' views on the development of the data market.

At the same time, we can also clearly see that Irys's narrative still leans towards B2B, with its value being more reflected in enterprise-level long-term storage and retrieval rather than personal data assetization. This is also one of the uncertainties in the current data market: compared to the demand for large-scale data from the B-side, personal data has a very low ceiling in terms of both value and impact.

Therefore, Irys's positioning itself as a supporting layer is reasonable: it avoids the uncertainty of personal data incentive mechanisms and focuses on the more stable demands of enterprises. However, this choice also means that in communities that narrate personal data sovereignty and in the Web3 ecosystem where personal participation is the core driving force, Irys's story inherently lacks direct communicative tension.

In this sense, Irys is not a representative answer to DataFi, but rather a supplement within the context of DataFi. It demonstrates another possible path from "storage" to "assetization," but whether it can be validated by the market still requires time and application to prove.