Just updated ComfyUI and saw HappyHorse 1.1 dropped as a Partner Node. The official positioning is straight-up production-ready—short dramas, e-commerce ads, brand content, game cutscenes, they name-dropped all of 'em.
For someone like me doing short-form video, the biggest win is that it renders dialogue, sound effects, and background music all in one go—no more running a separate dubbing pass. 1.1 mainly patched a few things: motion’s way smoother than 1.0, not so stiff anymore; multi-image reference to video can take up to 9 reference images; multiple characters keep their looks without face-swapping; you can even feed character and scene references separately, swap backgrounds without changing the person. Instruction following’s also stronger—they say it can handle prompts over 2500 characters, describing six to eight consecutive scenes in one go, auto-assigning duration and camera cuts. Skin texture’s been fixed for that oily, oversharp look, close-ups feel more human.
Three nodes—T2V/I2V/R2V—each doing their thing. Supports 720p and 1080p, 3 to 15 seconds, aspect ratios from 16:9 to 21:9, exports with built-in audio track. Honestly, having audio and video done in one shot saves a ton of post-production work. If it really handles long prompts like they claim, storyboarding short dramas with it will be way more comfortable. I usually throw complex multi-model jobs onto pixpix.com—one entry point with HappyHorse and other video models lined up, feed the same script in and compare which one syncs audio and video more solidly. Speeds up the pick.
