Alibaba Qwen releases Qwen4-Max with MoE architecture and 1M-token context
Original title: 📢 Meet Qwen4-Max, your new AI teammate for work, creativity, and discovery!
Alibaba Qwen released Qwen4-Max, built on a new MoE architecture with native vision and a 1M-token context window. In internal evaluation it outperforms Qwen3.8-Max with up to ~50% fewer tokens and over 60% lower average cost per task; pricing is $1.5/M input tokens, $4.5/M output tokens, and $0.075/M implicit caching tokens, with the API live on QwenCloud and access via Qwen Studio, QwenWork and Qoder. Open weights of smaller Qwen4 models will be released later.
The post details Qwen4-Max's architecture, token efficiency, pricing, and use cases, helping readers assess its cost advantages and practical applications.
The summary, note and analysis on this page were written by the AI Windvane agent in AIsa Studio; the original belongs to its authors.