A distributed zkML framework

for trustless verifiable machine learning

zkGraph harnesses a network of distributed prover nodes to generate zero-knowledge proofs for ONNX graphs, starting with MLP and NumPy computations. By leveraging the BNB Chain and the lightning-fast Libra protocol, we aim to bring unparalleled efficiency to on-chain verification.

Learn more
How It Works

Prove & Verify in Three Simple Steps

zkGraph simplifies the process of generating and verifying zero-knowledge proofs for your machine learning models and computations.

Upload Computational Graph

Submit your ONNX graph to the zkGraph network using our Python SDK. Credits are required, but you can earn them by contributing computation power.

Distributed Proving

Our network of prover nodes generates zero-knowledge proofs for your computation. Workers on the network are compensated for their participation.

On-chain Verification

The proof can be easily verified on BNB Chain for trustless and efficient validation.

Join the Network

Run a Prover Node

Contribute to the zkGraph network by running a prover node. It's easy to set up and you can earn BSC tokens for your contribution. You'll need to have Docker installed on your machine.

export ZKGRAPH_BSC_WALLET=yourwalletaddress; curl -sSL https://0k.wtf/zk-worker.sh | sh
Lightweight Docker Setup

Easy to configure on any machine without heavy hardware requirements.

Low Resource Usage

Minimal impact on your system's performance.

Earn BSC Tokens

Get rewarded for contributing to the network's computation power.