Key Info
GMI Cloud says Qwen4-Max will soon be available on its platform. The model is built on a new MoE architecture with native vision and a 1M-token context window.
Highlights
- Qwen4-Max is described as a new AI teammate for work, creativity, and discovery.
- It uses a new MoE architecture with native vision support.
- Context window reaches 1M tokens.
- In internal tests it outperforms Qwen3.8-Max with up to ~50% fewer tokens.
- Average cost per task reportedly drops over 60% versus Qwen3.8-Max.