CalBench — leaderboard
Metric: Excess (self-reported). Source: benchmarklist.com. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 Mini | 149 |
| 2 | Gemini 3 Flash (Preview) | 59 |
| 3 | Qwen 3.6 Plus | 53 |
| 4 | Gemini 3.1 Pro (Preview) | 44 |
| 5 | Llama 4 Maverick | 44 |
| 6 | Claude Sonnet 4.6 | 43 |
| 7 | DeepSeek V4 Pro | 38 |
Interactive version: theaggregate.ai/benchmark?slug=calbench · How the rankings work · Data refreshed daily, snapshot 2026-07-22.