Alibaba Cloud has launched Qwen2.5-Omni-7B, a compact multimodal AI model capable of processing text, images, audio, and video inputs while generating real-time text and speech responses. The model can be deployed on mobile devices and is open-sourced on platforms like Hugging Face and GitHub. This release follows Alibaba's commitment to AI development, with CEO Eddie Wu emphasizing continued investment in expanding AI boundaries.