Show HN: Vev – Ask questions about a screenshot, get a probability per answer

Category: ai-ml

Tags: vision-language-model, decision-making, probability-estimation, screenshot-analysis, jev-api

Score: 8.0/10 (Innovation: 8, Technical: 8, Documentation: 9, Utility: 7)

Vev is an open-source vision-language model fine-tuned from Qwen3.5 that answers yes/no and multiple-choice questions about screenshots, text, or JSON, returning calibrated probabilities per answer in ~40ms. It implements TypeSafe's Jev API for decision-making, supports images, and ships with benchmarks, a serving CLI, Docker support, and Hugging Face weights.

Target audience: backend devs and AI engineers building decision models and image-aware automation

Repository: https://github.com/Xiaooolong/vev · Python · Apache-2.0 · 18 stars

View on Hacker News