Alibaba ATH AI & Zhang Di: The Team Behind HappyHorse 1.0

From Kuaishou to Alibaba: Zhang Di's Journey

The story of HappyHorse 1.0 begins with Zhang Di (张迪), one of the most influential figures in China's AI video generation space. As Vice President of Kuaishou — one of China's largest short-video platforms — Zhang Di led the development of the Kling video model, which became one of the most prominent AI video generators in the industry.

Zhang Di's departure from Kuaishou to join Alibaba was a significant event in China's AI industry. The move was reported by multiple major news outlets including the South China Morning Post (SCMP), CNBC, and Bloomberg, reflecting the high stakes of talent competition in the AI video generation space.

At Alibaba, Zhang Di was tasked with leading the ATH AI Innovation Unit — a specialized team focused on pushing the boundaries of AI video generation technology.

What Is Alibaba ATH AI?

ATH AI (also referred to as the ATH AI Innovation Unit) is a specialized division within Alibaba Group dedicated to AI video generation research and development. While Alibaba has multiple AI research divisions (including DAMO Academy and Tongyi Lab), ATH AI is specifically focused on video generation models.

Under Zhang Di's leadership, ATH AI developed HappyHorse 1.0 — which would go on to become the #1 ranked AI video model on the Artificial Analysis Video Arena in April 2026.

The Anonymous Debut

HappyHorse 1.0 first appeared on the Artificial Analysis Video Arena on April 7, 2026, submitted anonymously without any company attribution. The Artificial Analysis platform uses a blind evaluation methodology — users compare video outputs from different models without knowing which model produced which video.

Within days, HappyHorse rapidly climbed the leaderboard through community voting, eventually reaching the #1 position. The anonymous submission generated significant speculation across Reddit, Twitter/X, and AI community forums about who had built this model.

On April 10, 2026, Alibaba officially confirmed that HappyHorse 1.0 was developed by their ATH AI Innovation Unit. The revelation that the model was built by the former Kling lead added another layer to the story, highlighting the intense talent competition between China's tech giants in the AI video space.

Why This Matters for the AI Industry

The HappyHorse story illustrates several important trends in the AI video generation industry:

1. Talent Migration Drives Innovation

Zhang Di's move from Kuaishou to Alibaba demonstrates how talent migration between tech companies accelerates innovation. His experience building Kling directly informed HappyHorse's architecture, particularly in motion synthesis and multi-shot storytelling.

2. China's AI Video Race

With Alibaba (HappyHorse), ByteDance (Seedance), and Kuaishou (Kling) all competing for the top position, China's tech giants are investing heavily in AI video generation. This competition is producing rapid improvements in video quality, speed, and capabilities.

3. Unified Multimodal Models

HappyHorse 1.0's unified audio-video generation represents a significant architectural shift. Rather than using separate models for video and audio (as most competitors do), HappyHorse generates both in a single pass. This approach may become the standard for next-generation AI video models.

Timeline

What's Next for HappyHorse?

As of April 2026, HappyHorse 1.0 is not yet publicly available. Alibaba's ATH AI unit has indicated plans for broader access, but no specific launch date, pricing, or open-source plans have been confirmed.

The AI community is watching closely for announcements about public API access, model weights release, and commercial licensing terms. We will update this page as official information becomes available.