HAERAE-Vision - Business/Economics: leaderboard
Metric: Checklist Score (%; original under-specified queries, GPT-5-mini judge, mean of 3 runs). Source: arxiv.org. Saturation forecast: Around December 2026. 45 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 | 58.41 |
| 2 | Gemini 2.5 Pro | 51.09 |
| 3 | GPT-5 Mini | 50.19 |
| 4 | Gemini 2.5 Flash | 48.05 |
| 5 | Qwen 3 VL 235B A22B (Thinking) | 47.3 |
| 6 | Grok 4 | 44.44 |
| 7 | Qwen 3 VL 235B A22B Instruct | 43.16 |
| 8 | Qwen 3 VL 30B A3B Instruct | 42.02 |
| 9 | Qwen 3 VL 32B Instruct | 41.29 |
| 10 | Gemini 2.5 Flash Lite | 38.84 |
| 11 | Qwen 3 VL 30B A3B (Thinking) | 38.44 |
| 12 | Qwen 3 VL 32B (Thinking) | 38.21 |
| 13 | Qwen 3 VL 8B (Thinking) | 33.92 |
| 14 | Gemma 3 27B | 31.71 |
| 15 | Qwen 3 VL 4B Instruct | 31.04 |
Interactive version: theaggregate.ai/benchmark?slug=haerae-vision-business-economics · How It Works · Data refreshed daily, snapshot 2026-09-25.