Fhenix

  • Name: Fhenix
  • URL: https://www.fhenix.io/
  • Category: confidential-compute infrastructure / FHE coprocessor / privacy-preserving smart-contract tooling
  • Tags: ethereum-ecosystem
  • Summary: Fhenix is now better read as confidential-compute middleware for Ethereum apps than as another would-be L2. The live story is CoFHE: Solidity libraries, a client SDK, onchain task and access-control contracts, and offchain FHE execution and decryption services. That product note still matters, but the named operator boundary around the current threshold-network rollout, trusted-dealer assumptions, docs/SDK packaging, and decentralization timing now lives separately in fhenix.
  • What it does:
    • Provides a Solidity FHE library and client SDK so developers can encrypt inputs, operate on encrypted values, and manage decryption permits in EVM-compatible applications
    • Runs CoFHE, a modular off-chain computation layer that executes FHE operations while coordinating with on-chain contracts such as the Task Manager, ACL, and ciphertext registry
    • Supports privacy-oriented application patterns like shielded stablecoins, confidential token balances, protected governance flows, sealed-bid auctions, and Uniswap v4 hook-based encrypted trading logic
    • Uses a threshold-decryption network plus ZK-verified encrypted inputs to keep data encrypted through most of the computation lifecycle
    • Publishes active code and examples across contracts, SDKs, docs, and proof-of-concept applications, including Redact and shielded stablecoin demos
  • Key claims:
    • The current homepage says Fhenix has evolved from an FHE Layer 2 into a confidential DeFi infrastructure company centered on CoFHE, positioning privacy as an add-on for Ethereum apps rather than only a separate chain environment
    • CoFHE architecture docs describe a hybrid system of on-chain contracts plus off-chain services including FHEOS Server, Slim Listener, Result Processor, Threshold Network, and ZK Verifier
    • The docs say encrypted inputs are accompanied by zero-knowledge proofs of knowledge before entering contract workflows, and decryption access is managed through permits and threshold cryptography rather than public plaintext exposure
    • The public roadmap is unusually explicit about current trust assumptions: Fhenix says threshold-network parties are run by Fhenix, key generation still involves a trusted dealer, parts of the system remain centralized, the codebase is unaudited, and not all code is yet fully open-source
    • The GitHub organization shows recent activity across developer docs, contracts, SDKs, and example applications, suggesting the current center of gravity is developer tooling and middleware rather than a standalone consumer chain launch
  • Whitepaper: No current canonical standalone whitepaper or litepaper was found in the official materials reviewed during this pass. The most informative primary sources were the official site, CoFHE docs, roadmap pages, and active GitHub repositories; see ../whitepapers/fhenix-primary-sources-2026-04-27.md.
  • Sources:

Internal linkages

  • Company / operator boundary for the current training-wheels implementation, threshold-network rollout, and docs / SDK shell: fhenix.
  • Best full-stack confidential-execution read: oasis-sapphire.
  • Best rollup-shaped confidential-execution contrast: ten-protocol.
  • Useful company-side confidentiality middleware comparison: inco.

Protocol / company boundary

  • Keep this project note on the confidential-compute product surface: CoFHE architecture, FHE.sol, encrypted inputs and outputs, threshold decryption design, and the application patterns the stack enables.
  • Use fhenix when the sharper question is who currently runs the threshold-network parties, owns the trusted-dealer and coordinator assumptions, controls docs and SDK defaults, and decides when the system becomes more open, audited, or decentralized.
  • That split matters because how does CoFHE work? and whose rollout and trust assumptions are developers inheriting today? are related but not identical questions.

Control surface

  • The mechanism layer is still the important product cut: encrypted state, offchain FHE execution, decryption permits, and EVM-friendly integration.

  • But the live trust surface remains operator-shaped. Current threshold-network parties, key-generation assumptions, reencryption flow, centralized input storage, and the timing of audits or open-sourcing point upward to fhenix.

  • Read Fhenix here as confidentiality middleware with visible implementation tradeoffs, not as a finished decentralization story.

  • Last reviewed: 2026-08-12 UTC