Alibaba: wanx2.1-t2v-plus API

Wanx2.1-t2v-plus is a state-of-the-art text-to-video foundation model under the Wan-AI series. It utilizes a 14-billion parameter Diffusion Transformer (DiT) architecture, specifically optimized for high-fidelity spatial-temporal consistency. Compared to the "Turbo" version, the "Plus" variant leverages a much larger parameter count and a refined VAE (Variational Autoencoder) to capture complex real-world physics, intricate human movements, and cinematic lighting. It is the first video model to achieve native high-precision text rendering in both Chinese and English, making it highly effective for advertising, short-form storytelling, and professional creative workflows.

• Input: text
• Output: video
• Released: 2025-02-24
• Knowledge cutoff: 2024-12-31