What Is Fhenix? Encrypted Computation for Blockchain

What is Fhenix?

Fhenix is a blockchain R&D company building privacy-preserving infrastructure for Ethereum and EVM-compatible chains using Fully Homomorphic Encryption (FHE).

Its tools enable confidential computing and encrypted smart contracts that let applications process sensitive data while it stays encrypted end-to-end.

Core definition: Fhenix integrates FHE so smart contracts can compute directly on encrypted data without decryption, keeping data private throughout processing.

Key Products

1. CoFHE (FHE Coprocessor) – Fhenix off-chain encrypted computation service for EVM smart contracts
2. FHE Solidity Libraries – FHE.sol library with encrypted integer types and homomorphic ops
3. Threshold Decryption Network – Multi-party decryption service for authorized reveals
4. Cofhe.js Client Side Library – A TypeScript SDK that enables frontend apps to encrypt inputs, manage access control via permits, and unseal encrypted outputs.

Understanding FHE

What is Fully Homomorphic Encryption?

FHE is an encryption method that supports mathematical operations (addition, multiplication, composition) directly on ciphertexts.

The results remain encrypted; when decrypted by an authorized keyholder, they match the result as if computed on plaintext.

Key properties:

• Arbitrary encrypted computation
• Data encrypted at rest, in transit, and during execution
• Computing parties never see plaintext or decryption keys
• Lattice-based, post-quantum secure
• Homomorphic ops preserve ciphertext structure

Read What is FHE in details here

CoFHE: The FHE Coprocessor

CoFhe, an FHE enabled coprocessor, is a privacy tool for Ethereum and other EVM-compatible chains that allows computation directly on encrypted data. It means your smart contracts can handle sensitive information (like bids, votes, or user data) without ever exposing it to the network. CoFhe works off-chain, so it’s fast and scalable, and its stateless design makes integration as easy as adding a single solidity library.

Initial Integration: Single-line Solidity import

import "@fhenixprotocol/cofhe-contracts/FHE.sol";


Workflow:

1. Smart contract emits an FHE computation request
2. Off-chain aggregator detects it and calls FHE engine
3. CoFHE computes on ciphertext
4. (Optional) Threshold network decrypts for authorized parties
5. Contract receives the hash (pointer) of the encrypted result for use

Primary components:

• Task Manager: Acts as the gateway for all FHE operation requests, validating requests and managing permissions through the Access Control Layer (ACL).
• Aggregator: Coordinates request queues and manages communication between on-chain contracts and the off-chain execution environment.
• FHEOS Server: Executes the actual FHE operations on encrypted data and maintains the encrypted state.
• Threshold Network: A distributed system that securely handles decryption requests through multi-party computation, ensuring no single entity can access the decryption key.
• Ciphertext Registry: Maintains references to encrypted values and handles access control.

All operations, including decryption, are asynchronous. This requires proper logic

Decentralization Roadmap (Planned)

While Fhenix is pioneering confidential computing for Ethereum using Fully Homomorphic Encryption, its CoFHE architecture currently operates under a set of early-phase trust assumptions—akin to Vitalik’s “training wheels” model for rollups.

Fhenix is actively working toward removing centralization and trust dependencies by:

• Eliminating the Trusted Dealer through public key ceremonies and MPC-based key share creation
• Running ZK-Verifier and Threshold Network nodes in TEEs as an interim trust minimization step
• Using EigenLayer AVS and fraud proofs to validate off-chain computation results
• Migrating to decentralized data availability layers and modular on-chain integrity enforcement
• Auditing and open-sourcing the codebase to improve transparency and trust

This phased decentralization is designed to preserve performance and usability while gradually shifting trust away from Fhenix-controlled infrastructure toward cryptoeconomically secured, verifiable, and community-operated systems. A more in depth look at tradeoffs and future plans can be found here.

Why Blockchain Privacy Matters

On public chains, all state, transactions, and balances are transparent. This creates:

• Transaction tracing & address linkage
• Public balance inspection
• MEV exploitation
• Privacy regulation conflicts (e.g., GDPR)
• Competitive exposure of strategies

Fhenix’s solution:

• Encrypted contract state variables
• Encrypted user inputs
• End-to-end encrypted architecture for storage, transport, and computation

Use Cases

DeFi & Finance

• Encrypted limit/market orders (MEV protection)
• Private LP positions and strategies
• Confidential lending and DAO governance
• Encrypted payment flows

Enterprise & Regulated

• GDPR-compliant on-chain data
• Encrypted healthcare record computation
• Supply chain price logic privacy
• Selective regulatory disclosure

‍AI

• Encrypted inference & analytics
• Confidential model parameters

Developer Integration

• Language: Standard Solidity with euint types
• Frameworks: Works with Hardhat, Foundry, Truffle
• Networks: Live on Arbitrum Sepolia, Ethereum Sepolia, and soon to be other testnets.
• Code sample:

import "@fhenixprotocol/cofhe-contracts/FHE.sol";
contract PrivateCounter {
   euint32 private count;
   function increment() public {
       count = FHE.add(count, FHE.asEuint32(1));
       FHE.allowThis(count);
       FHE.allowSender(count);
   }
   function getCount() public view returns (euint32) {
       return count;
   }
}


Performance

• Threshold decryption throughput: 64,319 decryptions per second -- approximately 20,000× higher than the prior state of the art (3.18/sec)
• Single-operation latency: 8.48ms — approximately 37× lower than the prior state of the art (315.6ms)
• Gas model: on-chain cost limited to event emission and state updates; heavy operations run off-chain
• Source: Fhenix's ACM CCS 2025 Distinguished Paper (ePrint 2025/1781), measured in the online phase with four parties at 1ms ping. Research measurements, not production figures.

Funding & Ecosystem

• Seed: $7M (Multicoin, Collider Ventures, others)
• Series A: $15M (Hack VC lead; Amber Group, Primitive, GSR, Stake Capital, Dao5, etc.)
• Strategic: Tandem (Offchain Labs) for Arbitrum ecosystem integration; BIPROGY, strategic investor and enterprise partner (Japan; annual revenue over $2.6B)
• HQ: Tel Aviv, Israel
• Company Size: 30 people
• Founder: Dr. Guy Zyskind
• CEO: Guy Itzhaki

Fhenix FAQ

Q: Is Fhenix live?

A: Yes. CoFHE is live on three public testnets: Ethereum Sepolia, Arbitrum Sepolia, and Base Sepolia, with mainnet planned to follow. You can build and test confidential smart contracts today.

Q: Is it EVM-compatible?

A: Yes. Contracts are written in standard Solidity, use FHE.sol, and can deploy to any EVM chain.

Q: Can anyone decrypt?

A: Only if the contract explicitly allows them using FHE.allow(addr, ciphertext) or by using sealed outputs and permits.

Q: What happens if I call decrypt too early?

A: Use FHE.getDecryptResultSafe() to avoid reverts and check readiness before using a decrypted value.

Q: How do I encrypt data before sending to a contract?

A: Use cofhejs.encrypt() with the right Encryptable type like Encryptable.uint32(42).

Q: What is a permit?

A: A permit proves a user’s identity and authorizes unsealing of sealed outputs using their keypair.

Q: How does it differ from ZK?

A: ZK proves correctness of computation; FHE performs computation privately.

Q: Is FHE quantum-safe?

A: Yes. FHE is built on lattice-based cryptography, the same foundation behind NIST's post-quantum standards, so FHE-encrypted data resists quantum attacks by construction. Even data encrypted today stays protected against future quantum computers.

Q: Does it use TEEs?

A: No. Privacy is cryptographic, not hardware-based.

Q: How fast is the current CoFHE deployment?

A: CoFHE runs heavy computation off-chain, so on-chain cost scales with encrypted handle references rather than with the complexity of the encrypted computation.
On decryption performance, Fhenix's ACM CCS 2025 Distinguished Paper (ePrint 2025/1781) benchmarks the threshold decryption protocol at 64,319 decryptions per second in throughput and 8.48ms single-operation latency — approximately 20,000× higher throughput and 37× lower latency than the prior state of the art. These are research measurements under controlled conditions and describe the protocol, not current production throughput.‍

Q: Where to start?

A: cofhe-docs.fhenix.zone — includes install guide, contract examples, and SDK links.