OpenGradient

What OpenGradient does

OpenGradient is a research lab building an L1 network for open intelligence, enabling every AI agent toolcall, model inference, and API request to be secure and verifiable on a decentralized ledger. The OpenGradient Network powers high-performance verifiable computing for AI, allowing users to host models, run secure inference, and deploy agents on-chain. It aims to turn AI into user-owned infrastructure, providing portable, encrypted memory vaults, auditable and forkable models, and a share of the upside when data improves an agent. The platform emphasizes sovereign data ownership, trustless inference verification, and co-owned value creation.

Products

• OpenGradient NetworkThe network for full-stack, verifiable AI, powering high-performance verifiable computing for AI.
• Model HubA platform to explore thousands of AI models hosted on-chain.
• On-Chain AI SDKA software development kit to build verifiable on-chain AI agents, workflows, and apps, supporting ML and LLM inference, model management, and automated workflows.
• OpenGradient ChatPrivate chat infrastructure for AI.
• Digital TwinsAI digital replicas of influencers for conversational interaction.
• BitQuantA personal AI quantitative analyst.

Key capabilities

• EVM Compatible
• Decentralized AI model repository
• Verifiable AI Inferences
• zkML Proofs + TEE Attestations
• Portable, encrypted memory vault
• Auditable and forkable models
• On-chain compute and proofs
• Verifiable security
• One-click deployment of sovereign, open-source intelligence
• ML Inference & Workflows
• LLM Inference with tool usage and TEE execution
• Model Management
• Python SDK
• CLI tool (opengradient)
• Claude Code Plugin

Use cases

• Host models
• Run secure inference
• Deploy agents on-chain
• Build verifiable on-chain AI agents, workflows, and apps
• Optimizing AMM trading liquidity pools with on-chain ML models
• ML Risk Forecasting for Web3 applications
• Streaming chat completion with tool calling

AI approach

OpenGradient is building a decentralized network for verifiable AI computing, enabling secure inference, model hosting, and on-chain AI agents. They focus on user-owned infrastructure, auditable models, and on-chain proofs for transparency and security.

Tech named: zkML Proofs, TEE Attestations, EVM Compatible, on-chain ML models, AI memory, contextual safety, x402 payment protocol, LLM inference, ML inference

Industries served

• Web3

What it says sets it apart

• L1 network for open intelligence
• Enables every AI agent toolcall, model inference, API request to be secure and verifiable on a decentralized ledger
• User-owned AI infrastructure
• Portable, encrypted memory vault for users
• Models that users can audit and fork
• Users share in the upside when their data improves an agent
• Decentralized platform for user-owned intelligence
• Auditable open-source AI models
• Advancing the frontier of onchain cognition
• Sovereign data ownership
• Trustless inference verification on-chain
• Co-owned value creation where builders and users grow the network together

Funding rounds we track

From the AI funding tracker — rounds as reported by the linked publications.