ernie-5.0-thinking-latest

The next-generation Wenxin model, Wenxin 5.0, is a natively full-modal large-scale model that employs a unified, native full-modal modeling technique, enabling joint modeling of text, images, audio, and video, and equipping it with comprehensive multi-modal capabilities. Wenxin 5.0 features a comprehensive upgrade in its foundational abilities, delivering outstanding performance on benchmark test sets, particularly excelling in areas such as multi-modal understanding, instruction following, creative writing, factual accuracy, agent planning, and tool application.

ERNIE-5.0-Thinking-Preview boasts fully enhanced foundational capabilities, featuring stronger deep-thinking and visual comprehension skills, with exceptional performance in multi-modal understanding, instruction following, creative writing, factual accuracy, agent planning, and tool application.