Qwen3.8 Omni-Flash opens 1M-token multimodality

Alibaba has added Qwen3.8-Omni-Flash, a multimodal model that accepts text, images, audio, and video while exposing a 1M-token context window, to Model Studio, according to Alibaba Cloud’s Qwen3.8-Omni-Flash documentation. Qwen’s launch coverage reports more than 25% average improvement across 29 benchmarks and audio/video input costs cut by more than 93%, though those comparisons are vendor-reported in the release coverage. The archive’s Qwen3.8-Max preview pilot already treats Qwen switching as a workload-qualification problem: builders handling long media should test context fit and tool-call reliability, not infer a production win from the ceiling.