LexKairos - Bounded Event Retrieval: leaderboard

Metric: Accuracy (%; 594 items: retrieve the case events inside a time window, single- and multi-label; vanilla setting, the model answers directly without extra reasoning steps (thinking disabled where the model allows it); temperature 0, 8,000 output tokens). Source: arxiv.org. Saturation forecast: Around December 2026. 8 models tracked.

Top models

#ModelScore
1DeepSeek R185.69
2GPT-5.4 (Non-reasoning)84.68
3Gemini 3 Flash (Minimal)74.24
4DeepSeek V4 Flash (Non-reasoning)57.24
5Qwen 3.5 9B (Non-reasoning)36.36
6Llama 3.1 8B35.35

Interactive version: theaggregate.ai/benchmark?slug=lexkairos-bounded-event-retrieval · How It Works · Data refreshed daily, snapshot 2026-09-29.