IBA-Bench - Transportation: leaderboard

Metric: Checkpoint satisfaction (%; share of hidden preference checkpoints satisfied by the final action, transportation domain; mean of three LLM judges: DeepSeek-V3.2, Qwen3-4B-Instruct and Qwen3-30B-A3B-Instruct; retrieval-augmented baseline with bge-m3 top-3 chunks). Source: arxiv.org. Saturation forecast: Around January 2028. 10 models tracked.

Top models

#ModelScore
1Qwen 3 4B Instruct80
2QwQ-32B74.2
3GPT-5 Mini70.8
4DeepSeek V3.270.7
5GPT-5.170.6
6GPT-4o Mini68.3
7DeepSeek V367.3
8GPT-4o66.6
9Qwen 2.5 7B Instruct65.2
10GLM-4 9B Chat58.4

Interactive version: theaggregate.ai/benchmark?slug=iba-bench-transportation · How It Works · Data refreshed daily, snapshot 2026-09-26.