A comprehensive AI video generation skill for the LTX-2.3 API, enabling high-fidelity text-to-video, image animation, and precise audio-based lip-syncing.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install ltx-video
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install ltx-video using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is LTX-2.3 Video Generation?
The ltx-video skill is a sophisticated implementation of the LTX-2.3 video generation model, allowing developers to create cinematic content through a series of specialized API endpoints. As part of the Openclaw Skills ecosystem, this tool simplifies the complex process of generating AI video by handling direct MP4 downloads without the need for traditional polling. It excels at maintaining visual consistency and provides specific optimizations for both landscape and native vertical formats.
Whether you are looking to animate a static portrait or generate a complex scene from scratch, this skill provides the technical bridge to the LTX-2.3 pro and fast models. It is designed for high-performance applications where rapid generation times and professional output quality are paramount, making it an essential component for automated media workflows.
LTX-2.3 Video Generation Use Cases
- Generate high-quality cinematic videos from detailed text prompts.
- Animate still portraits into speaking videos using the audio-to-video lip-sync feature.
- Extend existing video clips to create longer narrative sequences or complex transitions.
- Retake specific segments of a video to fix visual artifacts or adjust character performance.
- Produce native vertical content for social platforms using resolutions optimized for portrait data.
How LTX-2.3 Video Generation Works
- The skill receives a request containing a prompt and optional media URIs (image or audio).
- It validates the input against LTX-2.3 technical requirements, such as HTTPS hosting and specific MIME types.
- A POST request is dispatched to the appropriate LTX-2.3 endpoint (text-to-video, image-to-video, etc.).
- The LTX-2.3 engine processes the request, taking between 17 and 60 seconds depending on the selected model.
- The skill receives the raw MP4 binary data directly from the API response.
- The resulting video file is saved to the local directory for immediate use in your Openclaw Skills workflow.
LTX-2.3 Video Generation Setup
To get started with this skill, you must have an active LTX API key. Save this key in your environment configuration:
# Add your API key to your environment
export LTX_API_KEY='your_ltx_api_key_here'
Ensure that any images or audio files you intend to use are hosted on an HTTPS server. You can use services like uguu.se for temporary hosting. For audio-to-video, ensure your audio is in MP3 format:
# Example upload command for media hosting
AUDIO_URL=$(curl -s -F "files[][email protected]" "https://uguu.se/upload" | python3 -c "import sys,json; print(json.load(sys.stdin)['files'][0]['url'])")
LTX-2.3 Video Generation Data Schema & Taxonomy
The skill manages data through structured JSON payloads and binary video outputs. The following table describes the primary data points used in LTX-2.3 Openclaw Skills operations:
| Parameter | Description | Required For |
|---|---|---|
prompt |
Textual description of movement, camera, and sound. | All Endpoints |
image_uri |
HTTPS URL of a source image (max 1080p). | image-to-video, audio-to-video |
audio_uri |
HTTPS URL of an MP3 audio file. | audio-to-video |
model |
Choose between ltx-2-3-fast or ltx-2-3-pro. |
All Endpoints |
resolution |
Target dimensions (e.g., 1920x1080 or 1080x1920). | All Endpoints |
duration |
Length of the generated clip in seconds. | Text/Image to Video |
LTX-2.3 Video Generation Advanced Features
- Native vertical support (1080x1920) trained specifically on portrait video data for superior social media content.
- Director-level scene blocking that respects spatial relationships like foreground/background and left/right movement.
- Advanced audio-to-video lip-syncing that maintains head levels and gaze lock for realistic character performance.
- Support for complex, multi-action prompting that allows characters to interact with objects and environments dynamically.
- Integration capabilities with ComfyUI through dedicated custom nodes for advanced node-based video pipelines.
- Video extension and retake functionality to chain clips together or refine specific portions of a sequence.
SKILL.md
Loading
Related Openclaw Skills
Brand Analyzer
An AI-driven research tool that uses Gemini and Google Search to generate structured brand identity profiles for creative automation.
Ad-Ready
Generate professional, brand-aware advertising images directly from product URLs using a multi-phase AI pipeline.
ElevenLabs Speech-to-Text
A high-performance transcription skill leveraging ElevenLabs Scribe for batch processing and real-time speech-to-text conversion.
Market Data Hub
A high-performance stock market data acquisition and technical analysis engine supporting multiple data sources and intelligent failover.
Nano Banana Pro
An advanced image generation and editing tool for Openclaw Skills that leverages Google's Gemini 3 Pro Image API for professional-grade visual content.
Morfeo Remotion Style
A specialized branding and development guide for creating Morfeo Academy vertical videos using Remotion and React.
Featured*
