Reference2Video allows users to upload multiple images, videos, and audio references to guide AI video generation. The platform controls subject appearance, motion, timing, style, and scene composition using models like Seedance 2.0, Kling 3.0, and others. It supports various aspect ratios, resolutions up to 720p, and video lengths from 4 to 15 seconds. The tool is designed for creators needing precise control beyond simple text-to-video generation.
Multi-Reference Video sits in PulseGate's AI & ML category. It focuses on creating consistent, reference-guided AI videos with precise control over subject motion, timing, and visual style. It is built as a consumer product for video creators and filmmakers. Multi-Reference Video follows a freemium model. It runs on the web.
It is developed by Reference2Video, and it first shipped in 2024. Among its 6 catalogued features are Multi-Reference Input, Subject Motion Control, and Style Transfer. The interface is available in German and English.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Other projects tracked alongside this one