FHE Companies (2026): Every Fully Homomorphic Encryption Company, Compared
The definitive directory of companies building fully homomorphic encryption products. Funding, team size, hardware requirements, performance benchmarks, and what they actually ship. Updated March 2026.
Fully homomorphic encryption allows computation on encrypted data without ever decrypting it. It's one of the most consequential advances in cryptography since public-key encryption. The industry has attracted over $2.1 billion in funding since 2021.
This directory covers every significant FHE company, what they've built, how much they've raised, and how their performance compares. We maintain this page as a resource for developers, CISOs, and investors evaluating the FHE landscape.
The FHE Companies
H33 — Post-Quantum Auth API
The only company shipping a complete post-quantum authentication pipeline as a single API call. Combines BFV fully homomorphic encryption, ZK-STARK proofs, CRYSTALS-Dilithium signatures, and ML threat agents. Encrypted biometric matching, fraud detection, KYC — all without decryption. NIST FIPS 203/204 compliant.
Hardware cost: ~$2/hr (single c8g.metal-48xl)
Zama — FHE Tooling & Infrastructure
The largest FHE-focused company by funding. Builds TFHE-rs (Rust library), Concrete (Python compiler), and fhEVM (encrypted blockchain). First FHE unicorn ($1B+ valuation). Series A $73M (Mar 2024), Series B $57M (Jun 2025), plus $44M token sale (Jan 2026). Backed by Multicoin Capital, Pantera Capital, Protocol Labs.
Hardware cost: $200-280K (8x H100 cluster) or $24-32/hr cloud
Duality Technologies — Enterprise Data Collaboration
Enterprise platform for privacy-preserving data collaboration. Built on OpenFHE, the most widely-used open-source FHE library. Focus on ML on encrypted data for healthcare, finance, and government. Awarded multimillion-dollar DARPA DPRIVE contract (TREBUCHET program) to build FHE hardware accelerator. Series A $19M, Series B $30M (Oct 2021). No new funding since.
Approach: Custom ASIC accelerator in development (DARPA-funded)
Enveil — Encrypted Search & Analytics
The only NIAP Common Criteria and CSfC-certified Data in Use security company. Focus on encrypted search and analytics for defense, intelligence, and financial services. Backed by In-Q-Tel (CIA), USAA, Mastercard, Capital One Ventures. Series B $25M (Apr 2022). No new funding reported in 2025-2026.
Differentiator: Government/defense certification (CSfC, Common Criteria)
Niobium Microsystems — FHE Hardware ASIC
Building purpose-built silicon for FHE acceleration. Samsung 8nm process, partnered with SEMIFIVE for ASIC development. Transitioning from prototype to production silicon. Targets 100-1000x acceleration over CPU-based FHE. Raised $23M (Dec 2025). Hardware not yet shipping.
Availability: Not shipping yet
Octra — General-Purpose FHE Network
General-purpose FHE network targeting encrypted token transfers and confidential DeFi. $4M pre-seed, $4M Echo round, $20M token sale at $200M valuation. Blockchain-focused, not enterprise API.
Mind Network — FHE Layer for Web3
FHE infrastructure layer for Web3. Building encrypted voting, FHE restaking, and confidential AI computation. $2.5M seed, $10M Pre-A (Sep 2024). Backed by Binance Labs, Chainlink, Hashkey Capital.
Fhenix — FHE on Ethereum L2
Building an Ethereum L2 rollup with native FHE support for encrypted smart contracts. Uses Zama's fhEVM under the hood. Strategic rounds from Offchain Labs (Arbitrum) and BIPROGY. Targets confidential DeFi and on-chain privacy.
Sunscreen — FHE Developer Tooling
Developer-focused FHE tooling. Builds the Secure Processing Framework (SPF), launched June 2025. Seed $4.65M (Jul 2022) led by Polychain Capital. Also runs a $500K grant program for FHE research. Coinbase Ventures, Northzone, dao5 also invested.
Lattica — FHE for Private AI
Early-stage startup focused on FHE for private AI computation in the cloud. Pre-seed $3.25M. Targeting encrypted ML inference where user data never leaves encryption.
Fabric Cryptography — VPU Chip for ZK + FHE
Building the VPU — a custom chip optimized for both ZK proof generation and FHE operations. Series A $33M (Aug 2024). Polygon purchased $5M of VPU systems. 60+ engineers from Nvidia, Apple, and Google. Targets mass production for hardware-accelerated cryptography.
Inpher — MPC/FHE Analytics (Acquired)
Multiparty computation and FHE for privacy-preserving analytics. Acquired by Arcium (Solana-based confidential computing) in November 2024. No longer independent. Previously backed by Amazon Alexa Fund, JPMorgan Chase, Swisscom Ventures.
Summary: FHE Companies Compared
| Company | Funding | FHE Scheme | Hardware | Best Latency | Status |
|---|---|---|---|---|---|
| H33 | Patent-pending | BFV (4 engines) | ARM CPU | 38.5 µs | Production |
| Zama | $150M+ | TFHE | H100 GPU | ~800 µs | Production (lib) |
| Duality | $49M | OpenFHE | CPU + ASIC (dev) | Undisclosed | Enterprise |
| Enveil | ~$40M | Proprietary HE | CPU | Undisclosed | Enterprise |
| Fabric | $39M | ZK + FHE | Custom VPU | Pre-production | Hardware dev |
| Octra | $28M | FHE network | Network | Undisclosed | Token launched |
| Inpher | $26M | MPC/FHE | CPU | N/A | Acquired |
| Niobium | $23M | FHE ASIC | Samsung 8nm | Pre-production | Hardware dev |
| Mind Network | $12.5M | FHE layer | Network | Undisclosed | Web3 |
| Fhenix | ~$7M | TFHE (Zama) | L2 rollup | Seconds | Testnet |
| Sunscreen | $5.2M | Compiler | N/A | Tooling | Developer tools |
| Lattica | $3.25M | FHE for AI | Cloud | Pre-product | Pre-seed |
Total industry funding: $2.1B+. Of the 12 companies listed, only 4 have production-grade FHE products (H33, Zama, Duality, Enveil). Two are building custom silicon that hasn't shipped yet (Niobium, Fabric). The rest are early-stage, blockchain-focused, or acquired. H33 is the only company running the full FHE + ZK + PQ pipeline on commodity ARM hardware at microsecond latency.
The Build vs. Buy Question
Every company on this list — except H33 — is building FHE as a technology for developers to integrate themselves. You get a library, a compiler, or (eventually) a chip. You still need to hire cryptographers, design your own protocol, manage GPU infrastructure, and maintain the system.
H33 is the only company that ships FHE as a finished product: one API call, complete post-quantum authentication pipeline, 38.5 microseconds, no GPU, no cryptography team required.
For a deeper analysis of the economics, see Build vs Buy Post-Quantum Encryption: Why $2.1 Billion in FHE Startups Can't Beat One API Call.
Skip the GPU cluster. Ship post-quantum encryption today.
1,000 free auths. Full FHE + ZK + Dilithium pipeline. No credit card required.
Get Free API Key →Methodology: Company data sourced from Crunchbase, PitchBook, company press releases, SEC filings, and published benchmarks as of March 2026. Performance numbers are self-reported by each company. H33 benchmarks are independently reproducible — see full benchmark suite. If you represent a company listed here and believe any information is inaccurate, contact support@h33.ai and we'll update it within 24 hours.