Keevx Image To Video

What This Skill Does

The Keevx Image-To-Video skill is a powerful integration designed for the OpenClaw AI agent to bridge the gap between static imagery and dynamic video content. By leveraging the advanced Keevx API, this skill empowers users to transform single images or collections of reference images into high-quality video sequences. It handles the entire lifecycle of video generation, including uploading source files, selecting between specialized generation models (Model V and Model KL), configuring aspect ratios, and defining resolution and duration preferences. The skill acts as an intelligent intermediary, automatically determining the best model based on user requirements—such as whether audio is necessary (Model V) or if the project requires a multi-angle product showcase using several reference images (Model KL).

Installation

To install this skill, use the following command in your OpenClaw environment: clawhub install openclaw/skills/skills/baidu-xiling/keevx-image-to-video After installation, you must configure your API credentials by setting the KEEVX_API_KEY environment variable. Ensure this key is exported in your shell or configuration file to allow the skill to authenticate requests with the Keevx servers. You can acquire this key at https://www.keevx.com/main/home.

Use Cases

1. Dynamic Content Creation: Convert marketing photos into short, engaging video clips for social media advertisements.
2. Product Showcasing: Use Model KL to ingest multiple angles of a product and generate a cohesive, professional 360-degree style presentation.
3. Automated Editing: Incorporate the skill into a batch pipeline to automatically generate video variations from a single source image for A/B testing.
4. Cinematic Enhancement: Utilize Model V to generate high-resolution 4K video with integrated audio from a single high-quality photograph.

Example Prompts

1. "Convert this landscape photo into a 4K, 8-second cinematic video with audio: [image_url]"
2. "Create a product showcase video using these three images of my watch, focusing on the refined details."
3. "Generate a 1080p, 5-second video based on these 4 reference angles of my new furniture design."

Tips & Limitations

• Model Selection: Always use Model V for 4K requirements and audio-sync tasks. Use Model KL specifically when you have 2 to 7 reference images to describe a subject from multiple viewpoints.
• File Handling: If providing local files, the skill handles the upload to Keevx automatically before starting the task. Ensure images are under 20MB.
• Limitations: Model KL does not support audio generation. Keep prompt descriptions concise yet descriptive (maximum 1000 characters) to ensure the AI interprets your motion intent correctly.