DV-World - DV-Sheet: leaderboard
Native spreadsheet visualization on DV-World (2026): agents create and repair spreadsheet charts and build dashboards, run in the DV-World-Agent framework, with expert rubrics applied by an MLLM judge. Overall DV-Sheet score in percent; higher is better.
Metric: DV-Sheet score (%): the paper's overall over native spreadsheet chart creation, chart repair and dashboard tasks, run in the DV-World-Agent framework (data manipulation, multimodal perception and proactive interaction tools); expert rubrics applied by a Gemini-2.5-Flash judge where rubric-scored, means over evaluation trials; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 11 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro (Preview) | 40.48 | #119 |
| 2 | GPT-5.2 | 37.24 | #185 |
| 3 | DeepSeek V3.2 | 33.12 | #419 |
| 4 | Kimi K2 (Thinking) | 32.52 | |
| 5 | GLM-4.7 | 31.53 | #380 |
| 6 | GPT-4.1 | 31.27 | #505 |
| 7 | Grok 4 | 30.75 | #319 |
| 8 | Gemini 2.5 Pro | 30.73 | #276 |
| 9 | Qwen 3 Coder Plus | 19.87 | #720 |
| 10 | Qwen 3 235B A22B | 17.89 | #628 |
| 11 | Qwen 3 8B | 6.91 | #1873 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=dv-world-dv-sheet · How It Works · Data refreshed daily, snapshot 2026-11-21.