Each model was given the same set of writing prompts with no system instructions. The output was scored using the slopwash analyzer. Higher scores mean fewer AI writing patterns detected.

Placeholder data. Real benchmark results will be added once standardized test outputs are collected and scored.

Methodology: each model receives identical prompts with default parameters and no system instructions. Outputs are scored by the slopwash heuristic analyzer. Scores reflect AI pattern density, not writing quality.