Skip to main content
For routing decisions, use tier_verdicts. The fields below support logging, reviewer interfaces, and certification records.

Scores

FieldDescription
truth_scoreOverall AI probability. 0 = human, 1 = AI. Use for displaying confidence to reviewers
geo_scoreSecondary corroborating score, stable across audio processing variations
proto_score_v1Drives the press-safe tier. Include in certification records
geometry_fused_scoreInternal score. Available for logging
head_ahead_kIndividual detection signal outputs. Each is 0–1, higher = more AI evidence. Use for logging or reviewer interfaces
embed_temporal_stdAcoustic consistency across the track. Available for logging
reliabilityConfidence in the result based on audio quality and duration. Below 0.6 — treat with extra caution

Origin attribution

FieldDescription
originPredicted generator: "suno", "udio", "elevenlabs", "human", "unknown"
origin_confidenceConfidence in the prediction (0–1)
Origin attribution is supplemental — use it for logging and analytics. It does not affect verdict or tier_verdicts.

Audit fields

FieldDescription
calibration_versionVersion of the model calibration that produced this result. Include in certification records
processed_atISO timestamp of when analysis completed