Show HN: JevBench, a reproducible benchmark for typed decision models

JevBench is a reproducible benchmark for evaluating typed decision models across intelligence, calibration, speed, and cost metrics, with each weighted at 25% in an overall composite score. The benchmark currently tracks 534 decisions per system and ranks models from various providers, with Jev 1.1 leading the rankings at a score of 74.4, followed by SemIf (Qwen3.5-4B) at 73.1. The platform allows customizable filtering by regional hosting, data confidentiality policies, and other parameters to help users select models based on their specific requirements.