Image & Video

Early Access : The Easy prompt engine. With 20+ million dialogue combinations, full present environments 44 Music genres +

The new ComfyUI node packs 107K words and 57 environment presets for cinematic AI image generation.

Deep Dive

Developer LTX2 has launched a massive overhaul of the Easy Prompt Engine, a custom node for the ComfyUI Stable Diffusion workflow. Dubbed LTX2EasyPrompt-LD, this tool is a complete rewrite, ballooning from 2,000 to 14,700 lines of code. Its core is a massive dialogue library containing 107,346 words, enabling over 20 million potential dialogue combinations. Each character action can be paired with 30 selectable dialogue options that dynamically adapt to the scene's context, including optional elements like swearing.

The update introduces two major systems for granular creative control. First, a music genre selector offers 44 distinct styles—from Jazz and Hip-hop to K-pop and Synthwave—each mapped to specific lyrical registers and vocal styles to influence the scene's mood. Second, 57 environment presets provide detailed world-building for prompts, ranging from iconic real-world locations like Times Square at night to performance spaces like a K-pop arena and natural settings like an Amazon jungle. The developer notes this is a feature-complete early access release focused on refinement, with updates planned every two days to squash bugs.

Key Points
  • Massive 107,346-word dialogue library enables over 20 million contextual character dialogue combinations.
  • Includes 44 music genres and 57 detailed environment presets for precise control over scene mood and setting.
  • A complete rewrite for ComfyUI, growing from 2K to 14.7K lines of code, requiring a fresh install from GitHub.

Why It Matters

Dramatically reduces prompt engineering time for AI artists, enabling rapid creation of complex, cinematic scenes with coherent narratives.