Wan 2.2 A14B

• 🏢 Alibaba's text-to-video model from the open Wan family
• 🆕 Released in 2025 within Alibaba's Wan lineup
• 👁️ Generates video clips directly from text prompts
• 🎯 Aims for subject coherence across frames
• 💬 Served through Alibaba Cloud Model Studio video APIs
• 📏 Offered at multiple output resolutions
• 🌐 Part of a family spanning text, image, and video editing
• 🔧 Complements Alibaba's broader Qwen and Wan tooling

Alibaba

Alibaba Group is a Chinese multinational technology company founded in 1999 and headquartered in Hangzhou, Zhejiang. Originally built around e-commerce and cloud computing, Alibaba has become one of the most prolific contributors to open-weight AI research,…

Wan 2.2 A14B is a text-to-video model from Alibaba's Wan family, released in 2025. It generates video clips from text prompts and, per Alibaba's catalog description, targets subject coherence across frames. Alibaba serves Wan video models through its Model Studio platform, with text-to-video and related generation available via the DashScope-based APIs.

Within the broader Wan lineup, Wan 2.2 A14B sits below later releases such as Wan 2.5 Preview, Wan 2.6, and Wan 2.7, which continue the line toward higher-fidelity and more multimodal generation. Each successive version adds capabilities across the text-to-video, image-to-video, reference, and editing variants that make up the family.

As one of Alibaba's earlier Wan video entries, Wan 2.2 A14B established the naming and tiering pattern carried into later generations. It complements Alibaba's cloud-served Wan endpoints and the wider lineup of Qwen text, image, and editing models, giving users a consistent toolset across modalities. For exact resolution options, configuration, and usage limits, the Alibaba Cloud Model Studio documentation is the authoritative reference.