> For the complete documentation index, see [llms.txt](https://docs.fluxpointstudios.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.fluxpointstudios.com/orynq/proof-of-inference.md).

# Overview

**Orynq** is Flux Point Studios' verifiable-AI stack. It produces cryptographic proof that AI systems did what they claimed — signed receipts for AI behavior, threshold-certified by the [Materios blockchain](/materios-partner-chain/materios.md) and immutably anchored to Cardano L1.

> **Orynq supersedes Proof-of-Inference (PoI).** The earlier $AGENT-denominated PoI protocol — job fees in $AGENT, staker fee-shares, inference-pool bonds, and PoI governance — is retired and will not ship. $AGENT itself is being consolidated into cMATRA under the [cMATRA Token Merger](/materios-partner-chain/cmatra-token-merger.md). On Materios, Orynq services are paid as ordinary network fees (MATRA / MOTRA capacity), and network security is compensated with validator and attestor block rewards — see the [PoI Litepaper (Legacy)](/orynq/litepaper.md) page for the full mapping.

## Overview

Orynq addresses the fundamental challenge of AI transparency: how do you prove what an AI did, with which inputs, producing which outputs? The stack combines:

* **Cryptographic Receipts**: Every observation or process trace generates a signed receipt with deterministic, content-addressed hashes
* **Threshold Certification**: A committee of independent attesters reaches quorum on Materios before a receipt batch is certified
* **On-Chain Anchoring**: Compact fingerprints are immutably anchored to Cardano (metadata label 8746)
* **Independent Verification**: Anyone can verify the full chain of custody from receipt fields back to a Cardano transaction, without trusting any single party

## Components

<table data-view="cards"><thead><tr><th></th><th></th></tr></thead><tbody><tr><td><strong>Orynq SDK</strong></td><td>Developer toolkit for anchoring AI process traces via Materios to Cardano</td></tr><tr><td><strong>Orynq Observe</strong></td><td>SDK for publishing cryptographically-attested AI model observation receipts to a public registry</td></tr><tr><td><strong>Anchor-as-a-Service</strong></td><td>Managed API for anchoring AI process traces without running infrastructure</td></tr></tbody></table>

## Use Cases

### Enterprise AI Compliance

Create auditable records of AI decision-making for regulatory compliance and internal governance.

### AI Agent Verification

Prove that autonomous AI agents executed their tasks correctly and within defined parameters.

### Model Inference Auditing

Track which models processed which data, enabling accountability in AI pipelines.

### Research Reproducibility

Anchor experimental AI runs to enable independent verification of research results.

## Quick Links

* [Orynq SDK](/orynq/orynq-sdk.md) - Developer documentation
* [Orynq Observe](/orynq/observe.md) - Attested model observation receipts
* [Anchor-as-a-Service](https://fluxpointstudios.com/anchor) - Try the live demo
* [PoI Litepaper (Legacy)](/orynq/litepaper.md) - The superseded PoI-era specification, kept for the record
* [GitHub](https://github.com/Flux-Point-Studios/orynq-sdk) - Source code