Video Generator

Video Generator skill enables your Claude agent to create short AI-generated videos from text prompts, image inputs, or image sequences. Supported models include Runway Gen-3, Kling, and Luma Dream Machine — each with different strengths in motion quality, prompt adherence, and generation speed. The skill abstracts provider differences so you can switch models with a single config change.

Text-to-video generates 3–10 second clips from written descriptions: describe the scene, camera movement, lighting, and style, and the model renders the video. Image-to-video animates a static image — a product photo comes to life with subtle motion, a portrait blinks and breathes, a landscape scene plays with wind-moved trees. This mode is particularly popular for creating dynamic social media content from existing photos.

Camera control parameters let you specify movement: pan left, zoom in, dolly forward, static shot — giving you directorial control over the generated footage. Style presets (cinematic, anime, documentary, photorealistic) apply consistent aesthetic treatment across batches. Resolution options range from 720p (faster, cheaper) to 1080p for final-quality output.

Generation typically takes 2–5 minutes per clip depending on the model and length. The skill returns a download URL and local file path once complete. Cost per clip varies by provider and quality tier — budget approximately $0.05–$0.50 per second of generated video. Social media teams, marketers, and content creators use Video Generator to produce short-form video content for platforms like Instagram Reels, TikTok, and YouTube Shorts without video editing expertise.

Installation

clawhub install video-gen

Install: clawhub install video-gen