PPT-Eval - Easy: leaderboard
The percent of the 51 easy tasks of PPT-Eval (2026), a PowerPoint-for-the-web benchmark for computer-use agents, that an agent completes with a perfect task-specific rubric score, averaged over three runs. Higher is better.
Metric: Success rate (%) on the 51 easy 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 February 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.5 | 56 | #156 |
| 2 | Claude Sonnet 4 | 55 | #400 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=ppt-eval-easy · How It Works · Data refreshed daily, snapshot 2026-11-13.