Vev

Vev is an open-weight AI tool that extends Jev-like decision models by allowing image input alongside text. It runs locally on your GPU and provides probabilistic answers to yes/no, multiple-choice, or graded questions about text and images, useful for tasks like UI state reading, policy compliance checks, and generated image validation. It supports Python SDK integration and adheres to the Jev API format for compatibility with existing Jev code.