[TypeSafe AI Releases Jev 1.14, Adds Image and Bounding Box Features, Expands Context by 4x] TypeSafe AI has announced the release of the Jev 1.14 model, introducing image input, bounding box data types for object localization, and expanding context capacity to four times its previous size. Images can now be combined with text, and existing Jev models have also gained multimodal capabilities. According to the official statement, the new version includes comprehensive minor improvements, with multilingual support showing the most noticeable enhancement. The `jev-preview` now points to `jev-1.14.0`. The newly added features remain in experimental preview and have not yet reached production-grade standards.