Alibaba's AI video model rises to No. 2 in global rankings, as OpenAI's Sora and ByteDance's Seedance fall away

Alibaba's HappyHorse 1.1 AI video generation model has risen to the No. 2 position in global rankings, with a score of 1,444 in both text-to-video and image-to-video categories, outperforming Google's Veo-3.1 and xAI's Grok-Imagine-Video. The model's unified self-attention Transformer architecture processes text, image, video, and audio tokens in a single sequence, eliminating the need for separate models or post-processing tools. This architectural simplicity can reduce the total cost of ownership for enterprise buyers. The practical implication for engineers building AI systems is that they can leverage HappyHorse 1.1's API-first design and integration into enterprise software stacks to streamline video synthesis workflows.

⚡ Key Takeaways

• HappyHorse 1.1 scores 1,444 in both text-to-video and image-to-video categories, leading Google's Veo-3.1 by 69 points in text-to-video.
• The model's architecture is built around a 15-billion-parameter unified self-attention Transformer that processes all modalities in a single generation pass.
• HappyHorse 1.1 operates as a unified system, eliminating the need for third-party dubbing or post-processing audio tools, which can reduce total cost of ownership.
• The model is now live on Alibaba Cloud Model Studio with full API access for enterprise customers and developers.
• Alibaba is offering a 40% sitewide launch discount for the first two weeks, priced for volume, and backed by a $52.7 billion global infrastructure buildout.

The rise of HappyHorse 1.1 has significant implications for engineers building AI systems, as it provides a production-ready video synthesis solution that can be integrated into enterprise software stacks. The model's unified architecture and API-first design can streamline video synthesis workflows, reducing the total cost of ownership and improving efficiency.

✅ Practical Steps

1. Integrate HappyHorse 1.1 into enterprise software stacks using its API-first design.
2. Leverage the model's unified architecture to eliminate the need for separate models or post-processing tools.
3. Evaluate the total cost of ownership and potential cost savings of using HappyHorse 1.1.

Want the full story? Read the original article.