Qwen & Wan Model Families
Explore Alibaba’s AI Models - Qwen (LLM) and Wan (Media) - on Atlas Cloud.
Qwen Image Models
Qwen-Image is Alibaba’s open image generation model family. Built on advanced diffusion and Mixture-of-Experts design, it delivers cinematic quality, controllable styles, and efficient scaling, empowering developers and enterprises to create high-fidelity media with ease.
Wan2.5 Video Models
Wan 2.5 is Alibaba’s state-of-the-art multimodal video generation model, capable of producing high-fidelity, audio-synchronized videos from text or images. It delivers realistic motion, natural lighting, and strong prompt alignment across 480p to 1080p outputs—ideal for creative and production-grade workflows.
Wan2.2 Media Models
Wan 2.2 introduces a Mixture-of-Experts (MoE) architecture that enables greater capacity and finer motion control without higher inference cost, supporting both text-to-video and image-to-video generation with high visual fidelity, smooth motion, and cinematic realism optimized for real-world GPU deployment.
Qwen LLM Models
Qwen is Alibaba’s advanced multilingual LLM family, engineered for enterprise and developers to deliver powerful reasoning, code generation, and long-context understanding. Built for safety, customization, and seamless integration, it enables smarter, faster AI applications across any domain.
