Qwen3.8-Omni-Flash

Qwen's new-generation native omni-modal model, supporting 1M long context sequences, can directly accept text, image, audio, and video inputs, and is based on the Qwen3.8-Flash-Next architecture. The model is oriented toward Agent capabilities in real productivity scenarios: while possessing Agentic capabilities such as programming, text knowledge work, and GUI operation, it has achieved significant results in audio- and video-centric Agentic applications that require comprehensive processing of text, image, audio, and video workflows, such as video editing, music video creation, film and television production and commentary, audio/video-to-multimodal summaries, and audio/video conversations. It supports 2-channel and 4-channel spatial audio parsing, is compatible with DashScope and OpenAI protocols, and it is recommended to install the supporting Qwen-MM-Plugins to facilitate the integration of native multimodal capabilities by Agent frameworks.