Summary: Irys is best understood not as generic decentralized storage, but as a programmable datachain that tries to make large onchain data directly usable by smart contracts while also proving that data remains available over time. Its reusable mechanism is the combination of a multi-ledger storage pipeline, storage-oriented consensus and packing design, and an EVM-compatible execution layer that is meant to compute on stored data natively instead of treating storage as a detached archival service. That makes Irys a useful comparison class for Arweave-style permanence systems, EthStorage-style programmable storage, and newer claims that data availability and application execution should sit in one protocol surface.
What it does:
Positions itself as a blockchain that stores large volumes of data onchain, lets smart contracts use that data directly, and proves the data is still there over time
Routes data through a multi-ledger lifecycle rather than one flat storage plane, with a Submit Ledger for intake/validation and a Publish Ledger for finalized permanent data
Plans additional term-oriented ledger variants for different storage durations, making retention itself part of protocol design rather than an offchain service choice
Exposes an IrysVM execution layer so contracts can compute against stored data as part of the same protocol story
Uses storage-specific block, partition, packing, ingress-proof, fee, and payment-distribution machinery rather than only bolting storage onto a conventional smart-contract chain
Key claims:
The official homepage gives the clearest top-level classification: Irys is a blockchain that stores large volumes of data onchain, lets smart contracts use it directly, and proves it's still there over time
The official whitepaper title, Irys: A Programmable Datachain for Verifiable Storage and Onchain Computation, is analytically important because it makes the core claim explicit: the protocol is trying to join storage assurance and computation in one datachain design rather than merely offering cheap storage
The whitepaper table of contents shows the real mechanism surfaces are not just storage APIs but also partition lifecycle, multi-ledger handling, matrix packing, PoW/S consensus, a VDF for read-speed synchronization, ingress proofs, fee split logic, and Submit Ledger expiry/payment processing
The official docs/search snippets describe a two-stage ledger model where data first enters the Submit Ledger for validation and only later reaches the Publish Ledger for long-term verifiable storage
The docs’ ledger description says the Publish Ledger contains only data provably uploaded to the network through the Submit Ledger system, which matters because the protocol is explicitly trying to remove holes that would otherwise make onchain data computation less reliable
The official core-features page/search snippet says IrysVM is a high-performance, EVM-compatible execution layer designed to work natively with onchain data, so the project’s novelty claim is not just cheaper bytes but native storage-plus-compute coupling
The whitepaper contents also show term-fee and permanent-fee distribution as separate mechanism areas, which suggests retention class and payment timing are first-order protocol variables rather than incidental pricing details
Whitepaper: The official whitepaper has been saved locally as ../whitepapers/irys-whitepaper.pdf. See also ../whitepapers/irys-primary-sources-2026-05-09.md.