Who we are
About
Reads owns the read path of Ethereum state access: that users can read state, history and logs privately, verifiably and without depending on a single supplier.
The Reads team builds tools and solutions that protect what users are reading from Ethereum’s state and who is doing the reading, so that wallets, frontends, and dApps can query Ethereum without leaking metadata to servers.
The work spans PIR (Private Information Retrieval), network-level anonymization, and EIP-7864-ready execution-layer clients: for our PIR work, but also unlocking it for light clients and others looking to migrate to it today , and ecosystem activation through targeted grants and collaborations.
Collaborations
The Tor Project, collaborating hands-on on embedding Arti in the browser via WASM and scaling the WebRTC bridge.
Dr. Ling Ren and Yue Chen at UIUC, new double-stateless schemes; security and correctness analysis of the VIA scheme.
Prof. Jung Ho Ahn, Hyesung Ji, and Jongmin Kim from SNU, GPU acceleration of the insPIRe and VIA schemes; exploration of new schemes.
EF Stateless team, an EIP-7864-enabled Geth and the conversion of MPT to UBT for subsequent proving of equivalence of the binary-based shadow chain to the mainnet MPT-based chain, using Stateless Geth (Keeper) as a guest program in OpenVM.
Lambda, an EIP-7864-enabled Ethrex client and subsequent proof of equivalence to mainnet.
Rohit Narurkar: GPU-ready Rust implementation of the VIA scheme, notwithstanding the security issue being investigated.
If you work on privacy research or tools that align with the team’s work and approach, write to reads@ethereum.org. We’d welcome ideas for collaboration.
Talks
- Seminar: Applied Cryptography for Privacy in Ethereum (pdf): COSIC, KU Leuven · June 4 2026
- A practical path to private reads of Ethereum data (pdf): Eurocrypt, Workshop on Cryptographic Tools for Blockchains, Rome · May 9 2026
- Sharded PIR Design for Ethereum State, EthCC · Stateless Summit · 2026
- Network-level Privacy for Ethereum RPC Traffic, EthCC · Stateless Summit · 2026
01
What we own
three pillars, one per person
- 01
Origin
hides who is asking, and the metadata that gives it awayAnonymous transport for read traffic, so that IP, timing and volume stop identifying the caller. Standardised behind one interface, so which anonymity network carries the traffic stays the caller's choice.
- 02
Content
hides what was asked: address, slot, access patternPrivate retrieval at the sizes Ethereum state actually reaches, accelerated until it runs on hardware people already own, and measured openly on that hardware.
Sharded PIRPIR specificationsinsPIRe on GPUUniversal PIR interface
- 03
Correctness
hides nothing; it ensures the answer and its context are authenticProof-carrying responses a caller can check against consensus, and the execution-layer work that makes those proofs small enough for a light client to verify.
A fourth property, payment, is part of the interface a caller talks to, but it is not a pillar. We do not design the incentive mechanism, and a pillar with nobody standing on it is a wish. All four are in Foundations.
02
What we do not do
scope is a commitment
We do not run production RPC, indexing or bundler infrastructure.
We ship specifications, libraries and reference implementations, and drive the alliances that get them adopted. Operating a service would make us a competitor to the people we need as adopters.
We do not own state scalability or Ethereum data structures.
That belongs to the Protocol and State teams. We feed requirements back to them and build on what they ship.
We do not design new peer-to-peer networks.
We offer a curated discovery layer and optional privacy methods for fetching state at the edge, and integrate the anonymity networks that already exist.
We do not design the incentive mechanism.
What we are accountable for is that nothing we build stands between a working market and the people who would create one: payment signalled in the interface from the first version, no switching cost between suppliers, and public measurement of what serving reads privately actually costs.