zkml

Framework to prove inference of ML models blazingly fast

ezkl is an engine for doing inference for deep learning models and other computational graphs in a zk-snark (ZKML). Use it from Python, Javascript, or the command line.

Curated list of ZKML models and use cases made with Giza

Easy to use Verifiable AI and smart contracts interoperability.

Optimization for on-chain private machine learning.

Reversion strategy for Haiko AMM

Giza agents run provable ML models and act autonomously on Starknet using native Account Abstraction

Scoria AI is a decentralized AI agent framework on blockchain, enabling private, on-device Web3 intelligence for users and enterprises.

Distributed zkML

verifiable and privacy-preserving inference via sparse neural networks by ZK-SNARKs schema (halo2)

To leverage AIGC technology with NFT, marketplace and privacy.

A deliberate affront to egalitarian tokenomics—cryptographic proof that not all wallets are created equal. What can be claimed without reputation can be excluded without apology.

Build a decentralized AI infrastructure on Solana, enabling secure on-chain model training and creating a global marketplace for AI inference services.

Blockchain-based multi-agent AI system where agents autonomously collaborate, trade, and evolve via decentralized incentives.

Scoria AI is a decentralized AI agent framework on blockchain, enabling private, on-device Web3 intelligence for users and enterprises.

Sovereign MythOS: A decentralized OS blueprint for AI × Blockchain × Symbolic Recursion. Keys-to-proofs architecture with capability security, zkML, and recursive self-modification—provable, autonomous, and uncensorable. Built with OSS: CometBFT, ezkl, AnonCreds. 🜁 No gods, just sovereignty.

Build a decentralized AI infrastructure on Solana, enabling secure on-chain model training and creating a global marketplace for AI inference services.

Securing confidential data in database using ZKML based cryptographic approach with auto encoder based encoding

DeepProve is a framework that enables fast verification of neural network inference using zero-knowledge techniques. It supports various model types, proving their correctness efficiently while keeping data private. 🐙✨