Open Hallucination Index LogoOpen Hallucination Index
About Us

The Future of AI Verification

The Open Hallucination Index is the first independent, open-source platform for detecting and verifying AI hallucinations. We build trust in a world where AI-generated content is becoming ubiquitous.

Our Mission

Truth as an API Endpoint

Large Language Models have revolutionized how we interact with information. But with great power comes great responsibility: Up to 27% of all LLM outputs contain factual errors – so-called hallucinations.

We have made it our mission to close this trust gap. Not through censorship or restriction, but through transparent, traceable verification.

Claim Decomposition

Atomic Analysis

Knowledge Graph

Hybrid Verification

Citation Trace

Traceable

Real-time

<50ms Latency

Technology

How Verification Works

A multi-step process that combines linguistic analysis with knowledge graph technology.

01

Claim Decomposition

The Input Processor breaks down complex texts into atomic subject-predicate-object triplets. Each individual claim is isolated and analyzed independently.

02

Knowledge Graph Matching

The Verification Oracle matches triplets against a hybrid index: Trusted domains (government data, science) combined with semantic consensus graphs.

03

Scoring & Citation

Each claim receives a HallucinationScore (0.0–1.0) and a complete Citation Trace with direct links to confirming or refuting sources.

Our Values

Principles That Guide Us

Trust Through Transparency

Our entire methodology is openly documented. Every verification is traceable and based on verifiable sources.

Scientific Foundation

Our algorithms are based on peer-reviewed research and are continuously improved through academic insights.

Privacy First

Your data belongs to you. We don't store sensitive content and meet the highest GDPR standards.

Open Source Mission

We believe in democratizing AI security. Our core components are freely available.

Ready for Verified AI?

Get started with the Open Hallucination Index today and bring trust to your AI applications.