X · by keno
How to Give JEV Eyes
Jev takes text, so turn the picture into text first.
keno
@kenonews
JEV can’t see images. So how do you give it eyes? What’s possible? What can run locally? How far can you get for free? The answers (and the catches) are in the full breakdown below:
Sep 21, 2026 · 18 likesOpen on X
keno’s article on giving Jev eyes. Jev accepts text and structured state, not images, so the picture is preprocessed into text, objects or other useful details before the decision. It compares the routes, from local OCR and vision-language models such as Qwen3-VL to cloud vision and image embeddings, and notes what each one costs and where it breaks.
Open the source
