Qwen3.5 Omni - Qwen’s latest generation of fully omnimodal LLM

>**Qwen3.5-Omni** is Qwen’s latest generation of fully omnimodal LLM, supporting the understanding of text, images, audio, and audio-visual content. Both the Thinker and Talker in Qwen3.5-Omni adopt the Hybrid-Attention MoE. Qwen3.5-Omni series includes Instruct versions in three sizes: Plus, Fla