kev

Kev is a family of small decision models built on Qwen3.5 and Qwen3.8 that you can use with pretrained weights or train yourself. ^[1] It answers yes/no (called noul), multiple-choice (choice), and rating (score) questions about a text in one request. ^[2] For example, it can classify a support ticket into a department, decide whether it needs escalation, and rate customer frustration. ^[3]