Show HN: CharacterTest.app–Scientific character matching using Big Five and LLMs

Hi HN,

I’ve always felt that traditional "Which character are you" quizzes were too static—relying on rigid branching logic and simple scoring that often yields predictable, low-accuracy results.

To solve this, I built CharacterTest.app. Instead of simple keyword matching, it uses an LLM-powered engine to perform semantic personality mapping. We decode user responses into the Big Five (OCEAN) and MBTI dimensions, then calculate proximity within a high-dimensional trait database covering 2,000+ characters across 100+ universes.

Technical Stack:

Framework: Next.js (App Router) with SSR for sub-second page loads and SEO.

AI Logic: Custom prompting architecture to prevent "personality bleed" and ensure multi-language consistency (supporting EN, JP, KR).

Privacy: No signup required for basic tests. I believe in "value first, data later."

I’m particularly curious about your thoughts on the mapping accuracy and the latency of the AI matching process. Would love to hear your feedback!