RTBENCHv0.4
← Back to scoreboard

Anam

closed · proprietaryMabyduck persona (default voice)last tested 2026-04-30T17:30:00Z

endpoint · Mabyduck personas (default voice)

homepage · https://www.anam.ai

All six probes vs. field median

Values normalized to [0,1]. IFC normalized as 1 − |v − 1|.

Raw scores

ProbeValueUnit
GR — Generation Rate0.127ratio
XPD — Cross-Run Pixel Diversity0.041embedding-dist
AVC — Audio→Motion Causality (off-mouth)0.320pearson-r
LWA — Listening-Window Activation1.210ratio
EVD — Emotional-Valence Differentiation1.420fisher-ratio
IFC — Idle-Loop Compression0.340ratio-vs-human

Frame strips

Identical-input replays. Hashes computed by the harness.

support-package-lookup

✗ identical

Composite history

Disclosure

Tested through the Mabyduck persona endpoint. Per-pixel variation above codec floor is non-zero in the lower-face region (lipsync overlay) but collapses elsewhere.