Raelume
RaelumeDocs
Text

LLM Block

Full prompt writing workspace with AI agents to write, fix, edit, and enhance prompts based on your request.

Inputs
ImageLLMPromptVariator
Outputs
LLMImageVideo3DWorldsVariatorWebhook

Features

Agent-Assisted Prompt Writing

Use built-in LLM agents to generate prompts from scratch, then iterate fast.

Fix, Edit, and Enhance

Ask the agent to rewrite tone, improve structure, or strengthen detail based on your request.

Full Text Editor Experience

Work in a complete writing space for long prompts, structured instructions, and multi-step prompt drafts.

Media-Aware Prompting

Connect images to analyze visual context and produce better aligned prompts.

Works with Reusable Prompt Blocks

Pair LLM with Prompt blocks when you want one clean prompt source reused across multiple targets.

Pro Tips

  1. Start with intent, then ask the agent to expand into a production-ready prompt
  2. Use iterative commands like "shorter", "more cinematic", "add camera language", or "clean this for model input"
  3. Connect LLM to Image/Video/3D/Worlds blocks to quickly test revisions
  4. Move finalized prompt text into a Prompt block when you want reusable distribution

Getting Started

  1. Add an LLM block to your canvas
  2. Write your goal, then ask the agent to produce or improve a prompt
  3. Refine through follow-up instructions (fix, edit, enhance)
  4. Connect to target blocks to run generations, or copy finalized text into a Prompt block for reuse

LLM is for deep prompt authoring. Use Prompt block for lightweight reusable prompt distribution.

Connections

Inputs
ImageLLMPromptVariator
Outputs
LLMImageVideo3DWorldsVariatorWebhook