A stealth model called HappyHorse-1.0 climbed to No. 1 on the Artificial Analysis video arena before Alibaba confirmed it was behind the surge, ahead of ByteDance and Kuaishou.
2026-06-23 · by Moe Ameen
In early April 2026, an unfamiliar text-to-video model called HappyHorse-1.0 appeared on the Artificial Analysis video arena — a blind, head-to-head leaderboard where users vote on which clip better matches a prompt — without naming its maker. It climbed to the top of the ranking, leading the text-to-video board and placing at or near the top for image-to-video as well, ahead of established models from ByteDance and Kuaishou. The anonymous debut and fast rise drew industry attention before anyone knew who was behind it.
On April 10, 2026, Alibaba confirmed it built the model, according to reporting from Bloomberg, CNBC, and Caixin. The company tied HappyHorse to its in-house AI effort and said the model was in limited testing, with broader API access through Alibaba Cloud's Model Studio expected to follow. The result it surpassed included ByteDance's Seedance 2.0 and Kuaishou's Kling line — and marked a sharp jump for Alibaba, whose earlier Wan video models had ranked further down the same board.
The surge is part of a wider pattern: Chinese labs now hold most of the top positions on third-party AI video rankings, competing on quality while undercutting Western models on cost. Note that HappyHorse is a separate, hosted model from Alibaba's open-weight Wan series; exact version numbers, benchmark scores, and pricing shift as the models update, so treat any single figure as a snapshot and confirm against Alibaba Cloud's listings before relying on it.
A leaderboard-topping model like HappyHorse wins on one thing: the raw clip. What it hands you is a few seconds of silent video — not a captioned, on-brand post sized for each platform, and not the supporting content a launch needs around it. That last mile is the whole job, and it is what Kompozy does. Bring a generated clip in and Kompozy burns in branded captions, reframes it to each destination's aspect ratio, composites it with b-roll or a music bed into a Clipped Short or Marketing Short, and fans the same idea into a carousel, a quote card, and platform-native captions in your voice — then schedules and publishes the set across all nine connected platforms from one queue. The model makes the footage; Kompozy makes the posts.
It also generates video the clip model cannot. Where HappyHorse outputs a short scene, Kompozy produces persona and avatar video — a talking-head HeyGen short with auto-captions, an avatar composited into a brand-exact HyperFrames template, a listicle video over a portrait clip — so your recurring, on-brand format does not depend on which generator is ranked first this week. And there is a news play today: drop your take on the Chinese-AI-video surge into Kompozy as a source and it spins one point of view into a blog post, a carousel explainer, short captioned clips, and posts scheduled across your channels. Being early on a story like this is how a single take becomes a week of content.
It is an AI text-to-video model that appeared anonymously on the Artificial Analysis video leaderboard in early April 2026 and climbed to No. 1. On April 10, 2026, Alibaba confirmed it had built the model through its in-house AI effort, with hosted access expected via Alibaba Cloud Model Studio.
On the Artificial Analysis video arena it led the text-to-video ranking ahead of established models from rivals including ByteDance (Seedance) and Kuaishou (Kling). Exact rankings and scores shift as models update, so confirm the current board before quoting a position.
No. HappyHorse is a separate, hosted model that topped the leaderboard in April 2026. Wan is Alibaba's open-weight video line, which is also highly ranked but distinct. Treat their versions, scores, and pricing as separate and check Alibaba Cloud's listings for current details.
The model only generates the raw clip. Bring it into Kompozy to add branded captions, reframe it for each platform, composite it into a Clipped Short or Marketing Short, fan it into a carousel and captions in your voice, and schedule and publish across TikTok, Reels, YouTube Shorts, X, LinkedIn, and more from one queue.