Qwen3.8 Launch - Dense200: leaderboard
Dense200 is a manually collected set of 200 densely annotated images, averaging about 91 small boxes per image across 109 categories, for detecting and grounding many small objects in crowded scenes. Scores come from the perception and grounding rows of the Qwen3.8-Max launch post's multimodal table; percent, higher is better.
Metric: Score (%). Source: qwen.ai. Saturation forecast: Around 2029. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.8 Max | 87 | #34 |
| 2 | Gemini 3.1 Pro (Preview) | 69.7 | #64 |
| 3 | Qwen 3.7 Plus | 60.7 | #98 |
| 4 | GPT-5.6 Sol | 55.3 | #22 |
| 5 | Claude Fable 5 | 31.1 | #20 |
| 6 | Claude Opus 4.8 | 20.8 | #39 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=qwen3-8-launch-dense200 · How It Works · Data refreshed daily, snapshot 2026-11-05.