Alibaba's Happy Horse 1.1 generates videos from text, animates a single image, or builds a video from multiple reference images. Supports 720p and 1080p, 3-15 second durations, and five aspect ratios.
Frequently asked questions
What is happyhorse-1.1 and who made it?
happyhorse-1.1 is a multi-mode AI video generation model created by Alibaba and available via Replicate. It can generate videos from text prompts, animate a single still image, or synthesize video from up to 9 reference images, making it a versatile tool for creators, marketers, and developers.
What video resolutions does happyhorse-1.1 support?
happyhorse-1.1 supports two output resolutions: 720p and 1080p. Users can choose based on their quality requirements and budget. Higher resolution 1080p output is ideal for professional or commercial use, while 720p offers a faster and more cost-effective option.
What aspect ratios are available in happyhorse-1.1?
happyhorse-1.1 supports five aspect ratios: 16:9 (landscape), 9:16 (portrait/vertical), 1:1 (square), 4:3 (standard), and 3:4 (portrait standard). These options cover social media, broadcast, and e-commerce formats, giving creators flexibility for different platforms and use cases.
How long can videos be with happyhorse-1.1?
happyhorse-1.1 generates videos ranging from 3 to 15 seconds in integer increments. This range is suitable for short social media clips, product showcases, animated stills, and brief promotional content. Users specify the exact duration as a whole-number value within this range.
Can happyhorse-1.1 animate a single image?
Yes. happyhorse-1.1 includes an image-to-video mode that takes a single still image as the opening frame and extrapolates natural motion to create a short video. This is useful for animating product photos, portraits, or illustrations without needing multiple source images.
What image formats does happyhorse-1.1 accept as input?
happyhorse-1.1 accepts input images in JPG, PNG, BMP, and WEBP formats, with a maximum file size of 10MB per image. For reference-to-video mode, users can provide up to 9 reference images to guide the model in maintaining visual consistency across generated video frames.
What is reference-to-video mode in happyhorse-1.1?
Reference-to-video mode allows users to supply up to 9 reference images — such as product photos, character sheets, or scene references — and generate a cohesive video that preserves subjects and visual identity across frames. It is ideal for brand-consistent content in e-commerce and advertising.
How can I reproduce the same video output in happyhorse-1.1?
happyhorse-1.1 supports reproducible results through an integer seed value. By using the same seed along with identical input parameters, users can regenerate the exact same video output. This is useful for iterating on a specific result or ensuring consistency across multiple production runs.
