CAME-Bench - Large: leaderboard
Metric: Answer-set macro F1 (%; 61 questions on two trajectories averaging 408K tokens; full context, left-truncated to the model's window, gpt-4.1-mini judge). Source: arxiv.org. Saturation forecast: Around January 2028. 5 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-4.1 Mini | 21.3 |
| 2 | GPT-5 Mini (Medium) | 21.2 |
| 3 | GPT-4o Mini | 1.7 |
| 4 | Qwen 3 235B A22B 2507 Instruct | 1.6 |
| 5 | DeepSeek V3.1 | 0 |
Interactive version: theaggregate.ai/benchmark?slug=came-bench-large · How It Works · Data refreshed daily, snapshot 2026-09-26.