阿里通义7大模型霸榜全球开源前十，千问Qwen3-Omni登顶

Core Insights - Hugging Face, the world's largest AI open-source community, released a new model ranking on September 28, showcasing Alibaba's Tongyi models, with seven models making it to the global top ten [1] - The newly open-sourced multimodal model Qwen3-Omni topped the list, capable of processing text, images, audio, and video, mimicking human capabilities of "listening, speaking, and writing" [1] - Qwen3-Omni demonstrates strong audio and video capabilities while maintaining stable performance in single-modal text and image tasks, marking a first in achieving such training effects in the industry [1] - The model is expected to be deployable in various scenarios, including in-car systems, smart glasses, and mobile devices [1]