PPT-Eval - Hard: leaderboard

Metric: Success rate (%) on the 30 hard tasks of PPT-Eval: share of tasks with a perfect task-specific rubric score (1.0), the agent editing the deck in PowerPoint for the web through GUI actions from full-screen screenshots; rubrics judged with Claude-4-Sonnet; mean of three runs; higher is better. Source: arxiv.org. Saturation forecast: Around May 2027. 7 models tracked.

Top models

#ModelScore
1Claude Opus 4.535
2Claude Sonnet 417
3Qwen 3 VL 32B11

Interactive version: theaggregate.ai/benchmark?slug=ppt-eval-hard · How It Works · Data refreshed daily, snapshot 2026-09-29.