What It Is
Arweave is designed for durable public storage rather than mutable cloud objects. Its social contract is "pay once, store for a very long time." The user submits a transaction containing data or a pointer to bundled data. The network stores it and makes it retrievable through gateways.
Important concepts:
- Blockweave. Arweave's chain-like data structure, where miners prove access to previous data.
- Proof of Access / Succinct Proofs of Random Access. Mining is tied to access to historical data.
- Storage endowment. Part of the upload payment is allocated to subsidize long-term storage.
- Permaweb. Web-like applications hosted on immutable Arweave data.
- Bundlr / Irys-style bundling. Batching layer used by many apps to reduce friction and improve throughput.
Where It Fits
Arweave is strongest when immutability is a feature:
- Publications and public archives.
- NFT metadata and art.
- Legal or governance records.
- Software release artifacts.
- Static frontends and permaweb apps.
- Datasets that should outlive the original operator.
It is weaker for private, mutable, or deletion-sensitive content. Anything sensitive should be encrypted before upload, and even encrypted data should be treated as long-lived.
Arweave vs IPFS/Filecoin
| Property | Arweave | IPFS | Filecoin |
|---|---|---|---|
| Primary goal | Permanent archive | Content-addressed retrieval | Paid storage market |
| Payment model | Pay once | External pinning | Deals over time |
| Mutability | Immutable data; mutable pointers via layers | Mutable names via IPNS/DNSLink | Deals can renew or expire |
| Retrieval | Gateways and nodes | Peers and gateways | Providers, retrieval markets, gateways |
| Best for | Public artifacts | Distribution | Large retained datasets |
Arweave competes more with "permanent publication" than with S3. IPFS/Filecoin is more modular; Arweave is more opinionated.
Trust Model
You trust:
- The Arweave economic model to keep storage incentives above storage costs.
- Gateways for convenience, unless the client verifies data against transaction IDs and hashes.
- Upload tooling and bundlers to correctly submit data.
You do not trust:
- A single hosting account.
- A domain registrar.
- A specific CDN.
The most important caveat is that public archival storage is intentionally hard to erase. That is a feature for censorship resistance and a liability for personal data.
Trade-offs
Strengths
- Strong fit for immutable public records.
- Simple mental model: pay once, retrieve through transaction ID or gateway.
- Broad adoption in NFT metadata, permaweb apps, and public web archives.
- Content persistence is part of the protocol's economic design.
Limitations
- Mutable app state needs layers above Arweave.
- Private data requires client-side encryption and key custody.
- Gateway availability and indexing still matter for user experience.
- The storage endowment model depends on long-run economic assumptions.
- Deletion and legal compliance are structurally difficult.
Attack Surface
| Attack | Mitigation |
|---|---|
| Gateway censorship | Use multiple gateways; verify transaction IDs |
| Bad upload path | Verify transaction inclusion and content hash |
| Plaintext disclosure | Encrypt locally before upload |
| Economic underfunding | Use Arweave for public archives with realistic risk tolerance; keep secondary copies |
| Mutable-pointer hijack | Sign manifests and use hardware-backed keys |
Primary sources
- Arweave whitepaper
- Arweave protocol documentation
- Arweave yellow paper
- Trautwein et al.: Design and Evaluation of IPFS
- Arweave protocol implementation
For alternatives and their operating obligations, see Choosing Decentralized Storage and BitTorrent for Archive Distribution.
Page history
- Updated
- 2026-10-25
- Origin
- Compiled from the ungovernable.tech research notes.
- Improve this page
- Suggest an edit — issues and pull requests are welcome.