Kev: Open-Source Decision-First AI Model Runs Locally as Jev Alternative

Kev is an open-source project designed as a locally runnable alternative to Jev, an AI model built to make fast, structured decisions rather than generate text. Unlike Jev, which was not released as an open-source model, Kev allows developers to download, inspect, modify, and run it on their own hardware. A hands-on evaluation tested four Kev checkpoints — 0.5B, 0.6B, 0.8B, and 4B parameters — on Apple Silicon using a support-ticket workload, finding that larger models did not simply increase confidence but meaningfully shifted probability distributions, uncertainty, and latency. Kev takes a state and typed questions as input and returns structured decisions with associated probabilities, making it suited for application logic rather than conversational use. The project is available on GitHub and Hugging Face, with larger checkpoints such as Kev-8B and 9B planned for follow-up testing.