Creative Fabrica offers early access to Alibaba’s HappyHorse 1.1 AI video model, bringing a massive leap forward in creative control and video quality.
Creative Fabrica, a leading platform for digital creators, today announced it is among the first to provide early access to HappyHorse 1.1, the highly anticipated upgrade to Alibaba’s AI video generation model. Following the successful launch of HappyHorse 1.0 on April 27 – which quickly became a staple for short drama production, e-commerce advertising, brand marketing, and game CG – version 1.1 introduces a massive leap forward in creative control and production-grade quality.
This release comes as Creative Fabrica and Alibaba continue to strengthen their partnership for bringing the highest quality video generation to users everywhere. HappyHorse 1.1 systematically addresses the highest-frequency needs of content creators by optimizing visual expression, character consistency, instruction following, and audio integration.
Marketing Technology News: MarTech Interview with Theresa Pham, Head of Product @ Wayvia
Key Capabilities and Upgrades in HappyHorse 1.1
– Significantly Enhanced Dynamic Expression: By rebuilding its motion modeling and inter-frame temporal consistency, HappyHorse 1.1 delivers significantly smoother action continuity. This directly resolves previous issues where generated movements could feel stiff or slow.
– Advanced Multi-Image Reference-to-Video (R2V): Creators can now achieve unprecedented character and scene consistency. The model supports Multi-Character Reference, allowing multiple distinct character images to be uploaded and maintained within the same video without cross-contamination. It also enables Character x Scene Free Combination, allowing characters to seamlessly maintain their appearance across entirely different scene transitions.
Marketing Technology News: Idle data is as good as no data
– Long-Context Instruction Following: The model’s semantic retention has been upgraded to support complex user inputs exceeding 2,500 characters. With Multi-Scene Automatic Scheduling, a single prompt can autonomously orchestrate and time 6 to 8 consecutive scenes, handling complex character positioning and camera blocking with high accuracy.
– Cinematic Visual Quality: Responding to industry feedback, HappyHorse 1.1 eliminates “oily facial appearances” and “over-sharpening”. It introduces naturalized skin textures, realistic detail rendering (preserving pores and soft skin qualities), and high-fidelity capabilities for large facial close-ups, making it ideal for high-end advertising and narrative storytelling.
– Integrated Audio Capabilities: The update features an enhanced audio-video coordination engine. Dialogue delivery now features natural variations in speech rate, pauses, and emotional tone. Background music can be explicitly controlled or disabled via prompts, and the precision of audio-visual synchronization has been vastly improved to eliminate desynchronization.
