LLM Stats (CountBench): leaderboard
Counting benchmark testing the ability to accurately count objects, items, or occurrences in visual or textual inputs.
Metric: Score (%). Source: llm-stats.com. Saturation forecast: Estimated already saturated. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.5 27B | 97.8 | #187 |
| 2 | Qwen 3.5 35B A3B | 97.8 | #256 |
| 3 | Qwen 3.6 27B | 97.8 | #170 |
| 4 | Qwen 3.6 Plus | 97.6 | #120 |
| 5 | Qwen 3.5 122B A10B | 97 | #173 |
| 6 | Seed 1.8 | 96.3 | #139 |
| 7 | Seed 2.0 Mini | 95.5 | #276 |
| 8 | Qwen 3 VL 235B A22B (Thinking) | 93.7 | #238 (Qwen 3 VL 235B A22B) |
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-countbench · How It Works · Data refreshed daily, snapshot 2026-10-20.