DocOps (Codex, Skills): leaderboard
Metric: Task pass rate (%; 210 native-format document tasks in Excel, Word, PowerPoint and PDF (50 atomic edits, 40 compositional, 60 single-document and 60 cross-document workflows), passed only when the submitted artifact satisfies every structural, semantic and preservation predicate of a deterministic in-container verifier; run through Harbor; Codex harness with the Anthropic xlsx, docx, pptx and pdf skills; higher is better). Source: arxiv.org. Saturation forecast: Around December 2026. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.5 | 67.1 |
| 2 | GPT-5.4 | 66.2 |
| 3 | DeepSeek V4 Pro | 42.4 |
| 4 | Gemma 4 31B | 33.8 |
| 5 | Qwen 3.5 27B | 29.5 |
| 6 | Qwen 3.5 122B A10B | 28.1 |
| 7 | Qwen 3.6 35B A3B | 21.9 |
| 8 | Qwen 3.6 27B | 21 |
| 9 | Qwen 3.5 35B A3B | 17.6 |
| 10 | Qwen 3.5 9B | 8.6 |
Interactive version: theaggregate.ai/benchmark?slug=docops-codex-skills · How It Works · Data refreshed daily, snapshot 2026-09-29.