Summary: Sia is best understood not as generic decentralized storage, but as a renter-host contract network where storage is governed through collateralized file contracts, host pricing, and proof windows on its own blockchain. Its reusable mechanism is the combination of competitive host markets, blockchain-enforced storage contracts, cryptographic storage proofs, and collateral burn for failed service, with a newer v2 stack that further separates wallet, renter, and host roles into modular software. That makes Sia a useful comparison class for Filecoin-style proof-weighted storage markets, IPFS-style content routing, and newer hot-storage layers that promise cloud-like UX without exposing where the real storage-enforcement logic lives.
What it does:
Lets renters form onchain file contracts with independent storage hosts who set their own storage, bandwidth, contract-fee, and collateral terms
Requires hosts to lock collateral and submit storage proofs during a proof window in order to recover collateral and earned revenue
Uses host-side competition and renter-side scoring to decide which hosts win contracts rather than routing demand through one protocol-run allocator
Pays hosts in Siacoin while burning collateral and earned revenue if proofs fail or hosts lose availability during the contract lifecycle
Exposes a modular v2 software stack with renterd, hostd, and walletd instead of one monolithic node, making storage, wallet, and app integration surfaces more explicit
Adds v2 features such as Utreexo-based fast sync, RHP4 renter-host transfers, prepaid accounts, browser-native access via QUIC, and supporting services like indexd
Key claims:
The official docs define Sia as the world’s safest cloud storage by design, emphasizing a purely decentralized peer-to-peer network with no central access point or gatekeepers; that is the cleanest short classification signal
The hosting docs make clear that Sia’s core mechanism is the file contract: renters lock payment funds, hosts lock collateral, and payout depends on proving data possession during a proof window
Those same docs show the real control surface for hosts is not just raw disk capacity but pricing, collateral, uptime, bandwidth policy, and contract formation competitiveness, with renter-side scoring done independently by each renter
The original whitepaper, linked from the official Sia blog, formalized Sia around decentralized storage contracts between peers rather than around one protocol-owned storage layer or one global retrieval market
The official retrospective says Sia shifted away from its original proof-of-storage consensus path while keeping the storage-contract mechanism, which matters because today’s analytical center is contract enforcement and host economics rather than “storage consensus” as such
The v2 hardfork materials show the protocol is still evolving materially: Utreexo compresses state for lighter nodes, RHP4 changes renter-host coordination, and the modular app stack plus indexd reduce operational friction without removing the underlying contract model
The hosting docs also highlight a 3.9% Siafund fee on contract payout, which is analytically useful because it makes rent extraction part of the contract surface instead of a hidden offchain business model
Whitepaper: The original Sia whitepaper has been saved locally as ../whitepapers/sia-whitepaper.pdf. See also ../whitepapers/sia-primary-sources-2026-05-09.md.