Unified Audio Generation
Generate picture and sound in a single pass, keeping effects, music, dialogue, and action in sync.
The #1 ranked AI video model on Artificial Analysis is coming to OlphinAI. Unified audio generation, cinematic 1080p output, and roughly 10-second creation speed.
About the Model
HappyHorse 1.0, also known as 快乐马, is a next-generation AI video generator developed by Alibaba's ATH AI Innovation Unit and led by Zhang Di, the former Kuaishou VP and technical lead behind Kling.
The model reached a 1388 ELO score on Artificial Analysis and earned the top ranking through blind community voting, setting a new benchmark for prompt adherence, motion, and cinematic output.
Unlike traditional generators, HappyHorse is built as a unified multimodal model that produces video and audio together, aligning music, sound effects, dialogue, and visual action in one generation pass.
Capabilities
Key HappyHorse 1.0 capabilities we are preparing for the OlphinAI launch.
Generate picture and sound in a single pass, keeping effects, music, dialogue, and action in sync.
Create native 1080p clips with rich lighting, stronger color grading, and broadcast-ready detail.
Build multi-scene stories while preserving character identity, pacing, and narrative continuity.
Produce cinematic video in seconds with an industry-leading speed profile and reliable completion.
Move from photorealism to anime, cyberpunk, watercolor, editorial, and more creative aesthetics.
Start from text prompts or transform still images into natural motion with expressive video synthesis.
Model Snapshot
The original preview details stay here, now presented as clean data cards instead of an image panel.
Model family
Unified multimodal video
Output
Video + synchronized audio
Quality target
Native 1080p
Best for
Cinematic stories, ads, characters
Get Notified
Sign up for an OlphinAI account and we will notify you as soon as HappyHorse 1.0 is ready for creators, teams, and developers.
Sign Up — Get NotifiedFAQ
Not yet. We are preparing the integration and will update this page as soon as HappyHorse 1.0 becomes publicly available.
HappyHorse 1.0, also known as 快乐马, is a next-generation AI video model from Alibaba's ATH AI Innovation Unit. It has been ranked #1 on Artificial Analysis for text-to-video performance.
Pricing details will be announced at launch. OlphinAI uses credit-based pricing across video models, so HappyHorse will fit into the same creator workflow.
Yes. After integration, we plan to make HappyHorse available through the same API experience as the other OlphinAI generation models.
Coming Soon to OlphinAI
In the meantime, explore the AI video generators already available in the OlphinAI workspace.
Try Other Models