Qwen (千问) launched Qwen3.8-Omni-Flash, supporting tool use, programming, graphical-interface operation, and workflows combining text, images, audio, and video.

Qwen (千问) announced the launch of Qwen3.8-Omni-Flash, described as its next-generation native multimodal model. The announcement emphasizes agent capabilities in practical productivity tasks, extending beyond understanding multimodal content toward “planning tasks, calling tools and completing creative work” (规划任务、调用工具并完成创作).

Reported general capabilities include programming, text-based knowledge work and graphical user interface (GUI) operation. The announcement also describes expanded agentic applications centered on audio and video, requiring coordinated handling of text, images, audio and video.

Named workflows include video editing (视频剪辑), music-video creation (音乐视频创作), film and television production and commentary (影视制作与解说), audio/video-to-text-and-image summaries (音视频转图文摘要) and audio/video dialogue (音视频对话). These examples are attributed to the Qwen announcement relayed by Financial_Express.

The source claims significant results in those workflows but supplies no numerical benchmarks, comparison methodology or independent evaluations. It also does not specify model size, pricing, latency, deployment hardware or licensing. The supported conclusion is a model launch with announced capabilities, not a quantified performance lead.