Enter your birth details and fixed catalog rules will compare the calculated Bazi signals with 32 public archetypes. A type is returned only when the available birth-hour scenarios agree sufficiently; five evidence axes then summarize model consistency. BZTI is free and is not a validated personality or mental-health test.
BZTI is a free birth-data archetype tool, not a questionnaire or psychometric personality test. Fixed catalog rules classify calculated Bazi signals against 32 public types. If birth time is unknown, a type is returned only when the candidate-hour results agree sufficiently. Separately, five axes—expression, action, emotional display, authority stance, and thinking style—contribute to the aggregate model-consistency summary. The result does not measure your real personality, mental health, wealth, or relationships.
BZTI begins with birth date, time, place, and the selected calculation settings. OpenFate validates that input and calculates the Bazi chart before any archetype wording is generated.
A fixed classifier scores the calculated chart signals against the public 32-type catalog. Five evidence axes are evaluated separately to summarize classification quality and model consistency. The result is useful for forming questions, but it is not a psychometric measurement of who you are.
The archetype is downstream of calculated stems, elements, Ten God relationships, structures, interactions, and related chart signals. The browser does not supply a finished type for AI to approve.
Fixed rules compare Ten God, strength, structure, interaction, and related chart signals with the 32 public archetypes. The classifier, not AI, determines the selected type or leaves the result unresolved.
When birth time is unknown, BZTI compares classifications across 24 candidate civil hours and requires sufficient agreement before returning a public type. If the candidates diverge too much, it keeps the result unresolved instead of inventing certainty.
Expression, action, emotional display, authority stance, and thinking style summarize supporting and opposing evidence after classification. The result card shows an aggregate consistency summary, not the raw five-axis trace. This is not a measured probability that the type matches your real personality.
AI can make selected narrative fields easier to read, but it cannot replace the chosen type, evidence axes, scores, or consistency data. Compare the card with repeated real behavior, context, and feedback rather than treating a symbolic label as a fixed identity.
Common questions about the 32 BZTI types, birth-time uncertainty, model consistency, and how this differs from a validated personality test.
OpenFate validates the birth input and calculates the Bazi chart. Fixed catalog rules then classify the resulting Ten God, strength, structure, interaction, and related signals against 32 public types. Five personality-evidence axes are added afterward to summarize classification quality and model consistency; AI chooses neither the type nor the consistency data.
The public catalog contains 32 fixed archetypes. The system may withhold a resolved type when the available chart evidence or unknown-time comparison is not consistent enough.
No. BZTI is a lightweight, screenshot-ready archetype card. A full Bazi reading examines the Four Pillars, Day Master, Ten Gods, Five Elements, interactions, and timing layers in greater depth.
No. BZTI uses birth-chart signals rather than questionnaire responses, and it has not been psychometrically validated. Neither the type nor the consistency indicator should be treated as a diagnosis or objective personality score.
The system can compare classifications across possible civil birth hours. It requires sufficient agreement before returning a public type; otherwise it can keep the result unresolved instead of inventing certainty.
No. Deterministic code owns the type, evidence, consistency, and symbolic indices. AI may only rewrite bounded narrative fields such as the summary and portrait, and a deterministic fallback remains available if that wording fails.