China AI Native Industry Insights – 20260805 – Alibaba | Moonshot AI | MiniMax | more
Today’s digest spotlights major advancements from Alibaba and MiniMax in the field of multimodal AI. This theme is underscored by Alibaba’s newly launched Qwen3.8-Max, a formidable model boasting 2.4 trillion parameters, with open weights expected next week. Meanwhile, MiniMax has introduced H3, another open multimodal AI model, which adds to the flux of AI innovations focusing on broader capabilities. Moonshot AI contributes by launching the Kimi Slides feature in Kimi Work, leveraging the power of Kimi K3 for enhanced productivity. Discover more in Today’s China AI Native Industry Insights.
1. Alibaba Launches Qwen3.8-Max, a 2.4T-Parameter Multimodal AI Model, with Open Weights Coming Next Week
Alibaba has officially launched Qwen3.8-Max, its most capable model to date, featuring 2.4 trillion parameters and a context window of up to 1 million tokens. The model ranks fifth in Text Arena and second in Vision Arena, with advanced capabilities in autonomous coding, long-horizon planning, and native multimodal intelligence including vision as a continuous feedback loop. Qwen3.8-Max is now available via API on Alibaba Cloud Model Studio at $2.0 per million input tokens and $6.0 per million output tokens, and can also be accessed through QwenWork, Alibaba’s all-in-one workplace AI agent platform. Open weights for Qwen3.8-Max and the smaller Qwen3.8-27B are scheduled for release the following week.
Read more: https://www.alizila.com/alibaba-unveils-qwen3-8-max-most-capable-flagship-model-to-date/
Video Credit: @Alibaba_Qwen on X
2. Moonshot AI Launches Kimi Slides Feature in Kimi Work, Powered by Kimi K3
Moonshot AI has introduced Kimi Slides, a slide-building feature within its Kimi Work productivity suite, powered by the Kimi K3 model. The tool automates the full presentation creation process, including content structuring, research, cohesive design, polished charts, and SmartArt generation. Finished slides are editable and available for download. The launch was accompanied by a tutorial video demonstrating the workflow. Kimi Work’s Slides feature positions Moonshot AI’s offering as an end-to-end AI-driven productivity tool for presentation creation.
Read more: https://x.com/Kimi_Moonshot/status/2084245860339298423
Video Credit: @Kimi_Moonshot on X
3. MiniMax Releases H3, an Open Multimodal AI Model
MiniMax has announced H3, an open model designed to work across multiple tasks and modalities. The release positions H3 as a unified solution capable of handling diverse input and output types without being constrained to a single domain. As an open model, H3 is intended to be accessible to developers and researchers. The announcement signals MiniMax’s continued push into open-source, cross-modal AI development.
Read more: https://huggingface.co/MiniMaxAI/MiniMax-H3
Video Credit: @MiniMax_AI on X
That’s all for today’s China AI Native Industry Insights. Join us at AI Native Foundation Membership Dashboard for the latest insights on AI Native, or follow our linkedin account at AI Native Foundation and our twitter account at AINativeF.