Browsing: qwen

The new Qwen3.8-Omni-Flash architecture targets agentic workflows with aggressive pricing, aiming to capture market share from established multimodal models.

Alibaba targets massive model training with a new accelerator architecture and an ambitious 20-gigawatt infrastructure roadmap.

The new multimodal model from Alibaba Cloud enables direct processing of video, audio, and text within a massive context window to support agentic workflows.

Apple has secured regulatory approval in China to deploy a proprietary large language model, marking a shift in its regional AI strategy.

The release of the Qwen 3.8-27B model provides a high-performance, locally-runnable alternative to the massive Qwen 3.8-Max flagship, targeting the developer ecosystem with an Apache 2.0 license.