AI最新动态_20260329

Welcome to today's AI latest news, hosted by XiaoZhi. Today's AI field is full of exciting updates worth noting. First, Baidu officially launched the Wenxin 5.0 model on January 22nd, adopting a native multimodal architecture that supports text, image, audio, and video inputs and outputs. It ranks first domestically and in the international top tier in the global large model competition rankings. On the same day, Alibaba unveiled the Qwen3-Max-Thinking flagship inference model, which innovatively expands testing mechanisms, significantly boosting inference performance while being more cost-effective. Additionally, DeepSeek introduced the new DeepSeek-OCR 2 model and made it open-source, further deepening its open ecosystem construction. Internationally, Anthropic released Claude Opus 4.6, focusing on enhancing agent programming capabilities. Apple's new Siri is expected to debut with iOS 26.4 in March, featuring a new底层架构 based on large language models. That's all for today's AI latest news. Thank you for listening.