Zama

Zama develops Fully Homomorphic Encryption technology and today sells the Zama Protocol, a confidentiality layer that lets assets be issued, transferred, traded and managed on public blockchains with amounts and balances encrypted at all times, including during processing. The protocol runs natively on any existing L1 or L2 chain, keeps transactions publicly verifiable, and supports programmable compliance so that regulated institutions can operate onchain. A live app suite lets users shield USDC or USDT into confidential tokens, unshield them, send, earn yield in confidential vaults and swap with the amount, direction and price limit encrypted. The buyers are institutions and builders in real world asset tokenization, payments and payroll, decentralized finance and trading, and token distribution, where exposing counterparties, positions or salaries is unacceptable. Developers integrate through Solidity guides, a TypeScript SDK and a confidential token standard, and Zama cites deployments with Morpho, Bron, Raycash, GSR and Blockscout on Ethereum mainnet. The underlying cryptography is available as open source libraries: a Rust implementation of TFHE, Concrete, a Python compiler for TFHE, and Concrete ML, an open source framework for training and running machine learning models on encrypted data. Zama does not publish pricing. The libraries are free to use, protocol usage is tied to the network's own token economics, and institutional engagements are arranged through the contact page. Buyers evaluating Zama as an AI tool should note the shift in focus. The company's public site is now entirely about confidential onchain finance, and Concrete ML, the privacy preserving machine learning library, lives in the developer documentation rather than on the product pages.