​
Overview

Traditional blockchain networks treat AI models as external resources, accessed through oracles or off-chain services. This creates a disconnect between smart contracts and the AI models they interact with, limiting the potential for truly decentralized AI applications.

Ritual takes a fundamentally different approach by making AI models first-class citizens on the blockchain, enabling atomic interactions between smart contracts and models.

Just as smart contracts revolutionized digital coordination, enshrined AI models enable a new paradigm of programmatic AI interactions with cryptographic guarantees.

​
Key functionality

Native Model Representation

Models exist as on-chain objects with:

Verifiable provenance

Modular model authenticity

Direct smart contract integration

Atomic Contract-Model Interaction

Enables powerful primitives like:

Model inference as part of transaction execution

Conditional transactions based on model outputs

Model-driven state transitions

Decentralized Infrastructure

Built on Ritual’s core components:

Modular storage layers for model artifacts

Verifiable execution through ZK, TEEs, and more

Native model marketplaces to track and trade models

​
Current State

  • Model registration and verification
  • Basic model ownership primitives
  • Verifiable inference execution
  • Integration with Symphony for scaling

​
Use cases

DeFi + AI

Automated market making with ML models

Model-driven portfolio rebalancing

Risk assessment for undercollateralized lending

Governance

DAO proposal analysis

Automated policy enforcement

Sentiment-driven treasury management

Gaming

NPC behavior models

Dynamic content generation

Player behavior analysis

Creator Economy

Generative art models

Content moderation

Personalization engines

​
Getting Started

Get started with consuming models with our Classical ML and LLM Inference sidecars.

Our tutorials & case studies feature hands-on ways to use native on-chain AI models, tapping into our enshrined compute and robust primitives.

Modular Computational IntegrityVerifiable Provenance