At today's Volcano Engine Force Conference, ByteDance made a grand announcement, officially introducing Doubao Large Model 1.8 alongside the enhanced audio-video creation model, Seedance 1.5 pro. According to evaluations, the Doubao model family has ascended to the forefront of the industry, excelling in multimodal understanding and generation, as well as showcasing remarkable Agent capabilities. Doubao Large Model 1.8 has undergone meticulous optimization specifically for multimodal Agent scenarios. This optimization has led to substantial improvements in tool invocation, adherence to complex instructions, and OS Agent functionalities, thereby significantly boosting its proficiency in tackling intricate tasks. Seedance 1.5 pro, on the other hand, stands out with its cinematic narrative flair, having achieved notable advancements in audio-visual synchronization technology. Additionally, it supports multilingual dialogue and ensures precise alignment of lip movements with various dialects.