Jev-Omni

A decision classifier for text, images, audio and video. Give it a situation and a question with options; it returns a probability for each option — no generated text, one forward pass.

Input type

text image audio video

Image / audio / video file (not needed for text)

Situation

Question

Options, one per line

Probabilities

Examples

Situation	Question	Options, one per line	Input type

Best supported at ≤20 options; audio is capped at 30 s; video is sampled to 16 frames. Weights: akhilaaa3/Jev-Omni, Apache-2.0.