LLM Stats (VLMsAreBlind): leaderboard
LLM Stats mirror of VLMsAreBlind, probing visual-perception blind spots and brittle reasoning in vision-language models.
Metric: Score (%). Source: llm-stats.com. Saturation forecast: Estimated already saturated. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.5 35B A3B | 97 | #268 |
| 2 | Qwen 3.6 27B | 97 | #182 |
| 3 | Qwen 3.5 27B | 96.9 | #192 |
| 4 | Qwen 3.5 122B A10B | 96.7 | #184 |
| 5 | Seed 2.0 Mini | 93.1 | #279 |
| 6 | Seed 1.8 | 93 | #143 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=llm-stats-vlmsareblind · How It Works · Data refreshed daily, snapshot 2026-11-05.