LTX-2 Easy Prompt & Vision nodes generate cinematic video prompts locally in ComfyUI
Two free nodes turn plain English or any image into 500+ token, cinema-ready prompts for LTX-2 video generation.
Deep Dive
A developer released two final, free ComfyUI nodes for LTX-2 video generation. The Easy Prompt Node structures prompts with priority-first logic and auto-generates negative prompts. The Vision Describe Node uses a local Qwen2.5-VL model (3B or 7B) to analyze images. Both run fully locally with no API, handle explicit content, and manage VRAM smartly, letting users create complex video prompts from simple text or images.
Why It Matters
It democratizes high-quality AI video creation by automating the most complex part: writing effective, structured prompts for professional results.