Skip to content
歸藏· op7418·· 2026-10-15AI score26

JEV 1.14.0 Update: Multimodal Support and 262K Context

JEV 昨天更新了一个大版本 1.14.0

AI briefing

JEV has released major version 1.14.0, officially supporting multimodal capabilities. The official statement notes that the model was inherently multimodal, but the first version did not include support. A new box type has been added, allowing the model to return the position of targets in images, suitable for scenarios requiring quick decision-making such as image processing, video editing, puzzles, and image editing. The context for each request and response has been increased to 262K, and multilingual capabilities have also been significantly enhanced.

Open original x.com