onchain-ai
(2)
OpenGradient
OpenGradient is a New York-based decentralized infrastructure company building the compute layer for verifiable AI. Its network hosts, executes, and verifies machine-learning models on-chain, attaching cryptographic proofs to every inference so applications, agents, and blockchains can confirm exactly which model ran, on what input, and what it returned. The stack pairs an EVM-compatible chain with a heterogeneous GPU/TEE compute architecture, a decentralized Model Hub of 2,000+ models, and developer SDKs, positioning it as an open, auditable alternative to closed cloud AI endpoints.
Matthew Wang
Matthew Wang is the founder and CEO of OpenGradient, a New York company building a decentralized compute layer for verifiable and confidential AI. A former Two Sigma research engineer who cut his teeth on equity-options market-making infrastructure, he left quant finance in 2024 to chase a simpler question: can you prove that the AI answering you is the AI you think it is? OpenGradient runs models on-chain with cryptographic proofs and hardware enclaves, and has raised $9.5M led by a16z crypto, with backing from Coinbase Ventures, SV Angel, and angels including Balaji Srinivasan, Illia Polosukhin, and Sandeep Nailwal.