Observed arrival · 2026-11-08
TuckerBench tests the voice of Malcolm Tucker
An independent benchmark compares how well language models sustain Malcolm Tucker’s voice in single replies and longer conversations.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- Language-model evaluators and Malcolm Tucker fans
- Worth noticing
- Character fidelity carries 40% of the score; human calibration and uncertainty analysis are still pending.
Field notes
The rubric separates recognisable rhythm, precision, specificity, and menace from comic invention, natural delivery, conversational fit, and persistence. Two model judges score anonymised transcripts, and the page flags self-family assessments and differences in provider conditions. Its own account says human calibration and uncertainty analysis remain pending, while the captured rankings and sample area display loading messages.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
◎NicheUnusually specific use
One card from the complete issue