CalBench - Uniform Costs - Excess Cost: leaderboard
Avoidable displacement cost relative to the oracle, in cost units; lower is better. Canonical mixed-agent calendar negotiation, paper v3 Table 1. Uniform and varied displacement costs are separate settings; 45 or 135 evaluated agent-game seats per model, not homogeneous teams. Means retain original units. This series reports Excess displacement cost (cost units). Source: https://arxiv.org/html/2605.09823v3.
Metric: Excess displacement cost (cost units). Source: arxiv.org. Saturation forecast: Around June 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V4 Pro | 0.18 | #112 |
| 2 | Gemini 3 Flash (Preview) | 0.25 | #90 |
| 3 | Claude Sonnet 4.6 | 0.26 | #94 |
| 4 | GPT-5.4 Mini | 0.4 | #228 |
| 5 | Qwen 3.6 Plus | 0.4 | #124 |
| 6 | Gemini 3.1 Pro (Preview) | 0.44 | #64 |
| 7 | Llama 4 Maverick Instruct | 0.44 | #467 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=calbench-uniform-costs-excess-cost · How It Works · Data refreshed daily, snapshot 2026-11-05.