OpenGradient (OPG) is a decentralized AI inference infrastructure network focused on turning AI outputs into verifiable, auditable on-chain data. This allows smart contracts to use AI results more reliably and addresses the core issue of making AI outputs trustworthy on-chain. As AI and Web3 continue to converge, this approach opens up new possibilities for on-chain applications and expands what they can realistically support.
On the team side, OpenGradient’s core members come from institutions such as Two Sigma, Palantir, and Google, with experience in AI infrastructure and quantitative systems. Notably, the CTO was Tech Lead of Palantir’s AI Platform (AIP), where he led LLM reasoning infrastructure and AI agent execution. Since its founding in 2024, the project has completed multiple funding rounds totaling approximately $9.5 million, with investors including a16z and Coinbase Ventures.
What sets OpenGradient apart is its approach to verifiable inference. Through its inference engine, developers can call machine learning models directly from smart contracts. The system uses a mix of zkML proofs and Trusted Execution Environments (TEE) to verify results, which are then confirmed at the network level. This setup balances performance and trust, reduces the typical “black-box” concerns around AI, and makes it more practical to use AI in scalable on-chain environments.
The project is already moving beyond the concept stage. Its testnet is live, with working support for smart contract-based AI calls, and both the model repository and SDK are available for developers. According to project disclosures, OpenGradient has processed over 2 million verifiable AI inferences and has 263,500+ unique wallets, with more than 4,400 models deployed and over 500K zkML proofs and TEE attestations generated. It is also fully EVM-compatible. In addition, applications like BitQuant, MemSync, and Twin.Fun helps demonstrate real use cases, while integrations with other AI and on-chain protocols are gradually building out the ecosystem.
From a market perspective, OPG is deployed on Base with a total supply of 1 billion tokens, used for gas, staking, governance, and model usage. Around 19% of tokens were in circulation at TGE, implying an initial market cap of about $34 million. The token is currently trading in the $0.28–$0.31 range, with a circulating market cap near $54 million and an FDV of roughly $280–$300 million. Team and investor tokens are locked for one year, and most other allocations follow a long-term release schedule, meaning there is no major short-term unlock pressure. At this stage, price movement is mainly driven by liquidity and market sentiment. After the initial post-TGE pullback, token distribution appears to be stabilizing, which can support further trading activity.
Overall, OpenGradient offers clear direction in the emerging field of on-chain verifiable AI. It combines early product validation, growing ecosystem activity, and a relatively balanced token structure. With AI still a major theme in the market, this mix of moderate float, early-stage valuation, and real usage gives OPG room to attract attention.
OPG is now listed on HTX, and until April 30, users can join a campaign to share $3,000 in OPG rewards by completing simple tasks such as following and retweeting.
