Human originality is becoming harder to prove.
Hapax analyzes human-AI threads, using the Hapax Originality Index (HOI) to quantify human/AI conceptual contributions.
Upload a transcript of a human-AI collaboration
Supported formats: TXT · JSON · CSV · DOCX · PPTX
Originality Analysis
Proof of Intellectual Authorship
Conceptual Density Rating
Model Judge
Gemini 2.5 Flash (Google)
Verification Standard
Hapax Infrastructure v1.0
Hapax is built for enterprise and academic security. Transcripts are parsed entirely client-side. When analyzed, data is sent to the evaluation engine via secure, ephemeral serverless functions. We do not have a database. Your chat transcripts are never logged, stored, or retained by Hapax.
The core markets where Hapax builds trust and establishes provenance.
Universities struggle with AI-assisted essays, research co-creation, and thesis integrity.
In AI-assisted research: Who actually created the hypothesis that drove the experiment?
Companies will increasingly ask: “Did this idea originate from the employee or the AI?”
Hapax performs a structured analysis of human-AI conversations using an LLM-as-a-Judge evaluation framework. The system evaluates transcripts using a 4-part rubric that reflects how ideas and code typically emerge in collaborative work:
We are actively building the next iteration of Hapax to integrate seamlessly into your daily creative workflows and institutional infrastructure.
A public 'Agora' where creators can publish verifiable HOI certificates, showcase their prompts, and debate the mechanics of human-AI collaboration.
Providing academic institutions and AI labs with anonymized, opt-in data to study how humans prompt and where creativity happens in AI-augmented workflows.
Generate certificates that aggregate originality scores across multiple judging engines, including OpenAI, Anthropic, and Google models.
Integrate Hapax certification into institutional workflows for academic integrity, research attribution, and corporate IP auditing.
In linguistics, a hapax legomenon is a word that appears only once in an entire body of text.
As artificial intelligence scales, it trains on the median of human thought, threatening to homogenize creativity. True originality is becoming rare. We built this platform to ensure that in a sea of algorithmic generation, your unique conceptual spark—your hapax—is isolated, measured, and protected.