A comprehensive suite of AI-powered prompt enhancement nodes for ComfyUI using local LLMs (LM Studio or Ollama). Transform simple prompts into detailed, platform-optimized descriptions for video and image generation.

## 🎯 All Nodes (Under "Eric Prompt Enhancers" Category)

A comprehensive suite of 5 AI-powered prompt enhancement nodes for ComfyUI using local LLMs (LM Studio or Ollama). Transform simple prompts into detailed, platform-optimized descriptions for video and image generation.### 1. Video Prompt Expander

Simple video prompt expansion with presets and expansion tiers.

2. Video Prompt Expander (Advanced)

---Granular control over video aesthetics with detailed settings.

🎯 All Nodes### 3. Image-to-Video Prompt Expander

Vision model analyzes images and adds motion descriptions.

Find all nodes under the "Eric Prompt Enhancers" category in ComfyUI.

4. Image-to-Image Prompt Expander

1. 🎬 Video Prompt ExpanderPlatform-aware image-to-image prompt generation (Flux, SDXL, Hunyuan, Qwen, Wan).

Simple video prompt expansion with style presets and expansion tiers.

5 Expansion Tiers: Auto, Basic, Enhanced, Advanced, Cinematic### 5. Text-to-Image Prompt Enhancer ⭐ NEW
6 Style Presets: Cinematic, Surreal, Action, Stylized, Noir, RandomAdvanced multi-platform image prompt enhancement with extensive controls.
Auto-Detection: Analyzes input complexity and selects optimal tier

Supports: Flux, SDXL, Pony Diffusion, Illustrious XL, Chroma, Qwen Image, Qwen Edit, Wan Image

2. 🎬 Video Prompt Expander (Advanced)

Granular control over video aesthetics with 50+ detailed settings.## ✨ Key Features

Lighting Controls: Light source, lighting type, time of day
Camera Controls: Shot size, composition, lens, angle, movement### Video Nodes
Color/Style: Color tone, visual style, visual effects- 🎬 5 Expansion Tiers: Auto, Basic, Enhanced, Advanced, Cinematic
Motion/Emotion: Character emotion, movement dynamics- 🎨 6 Style Presets: Cinematic, Surreal, Action, Stylized, Noir, Random
🎯 Auto-Detection: Analyzes input complexity

3. 🖼️➡️🎬 Image-to-Video Prompt Expander- 🔄 Multiple Variations: Generate up to 3 variations

Vision model analyzes images and adds motion descriptions for video generation.- � Vision Support: Analyze images for video generation

Image Analysis: Automatic scene understanding
Motion Addition: AI-generated motion descriptions### Image Nodes (NEW!)
Style Integration: Combines visual analysis with expansion tiers- 🖼️ 8 Platform Support: Flux, SDXL, Pony, Illustrious, Chroma, Qwen, Wan
🎨 Advanced Controls: Camera, lighting, weather, time of day

4. 🖼️➡️🖼️ Image-to-Image Prompt Expander- � Wildcard Random: Auto-variety in batch generation

Platform-aware image-to-image prompt generation.- 📸 Reference Images: Optional 1-2 image inputs

5 Platforms: Flux Redux, SDXL Img2Img, Hunyuan Img2Img, Qwen Edit, Wan Edit- 🏷️ Platform-Specific: Automatic token optimization per platform
Transformation Controls: Style transfer, detail level, creativity
Vision Analysis: Understands source image characteristics### Universal
🤖 Local LLM Support: LM Studio and Ollama

5. ⭐ 📝➡️🖼️ Text-to-Image Prompt Enhancer (NEW v1.7)- 📝 Keyword Integration: Add LoRA triggers and custom terms

Advanced multi-platform image prompt enhancement with extensive creative controls.- � File Export: Save prompts with metadata

8 Platforms: Flux, SDXL, Pony Diffusion, Illustrious XL, Chroma, Qwen Image, Qwen Edit, Wan Image- ➖ Smart Negatives: Platform-optimized negative prompts
Reference Images: Optional 1-2 image inputs with visual analysis
Genre Styles: 22 styles (cinematic, horror, cyberpunk, etc.)## Installation
Prompt Length Control: Very Short to Very Long (20-400 tokens)
Subject Controls: Framing (14 options) and Pose (17 options)1. Navigate to your ComfyUI custom nodes directory:
Advanced Settings: Camera, lighting, weather, time, composition, color mood```bash
Special Syntax: Emphasis (keyword:1.5) and Alternation {a|b|c}cd ComfyUI/custom_nodes/


---

2. The node is already installed in: `video_prompter/`

## ✨ Key Features

3. Restart ComfyUI

### 🤖 Local LLM Support

- **LM Studio**: OpenAI-compatible API (recommended)4. All nodes will appear under: **Add Node → Eric Prompt Enhancers**

- **Ollama**: Simple CLI-based LLM server

- **Privacy**: All processing happens locally, no data sent to cloud services## Requirements



### 🎨 Platform-Specific Optimization### Python Dependencies

Each platform has unique requirements and prompting styles:```bash

pip install requests

| Platform | Style | Optimal Length | Specialization |```

|----------|-------|----------------|----------------|

| **Flux** | Natural language | 75-150 tokens | Photography, artistic |### LLM Backend Setup

| **SDXL** | Natural/tags hybrid | 40-75 tokens | Versatile, balanced |

| **Pony Diffusion** | Booru tags | Tag count | Anime, characters |**Option 1: LM Studio**

| **Illustrious XL** | Danbooru tags | Tag count | Detailed anime |1. Download from: https://lmstudio.ai/

| **Chroma/Meissonic** | Detailed natural | 100-200 tokens | Complex scenes |2. Load a model (recommended: Llama 3 or similar)

| **Qwen Image** | Technical descriptions | Medium | General purpose |3. Start the server (default: http://localhost:1234)

| **Qwen Edit** | Edit instructions | Medium | Image editing |

| **Wan Image** | Cinematography | 60-120 tokens | Professional video stills |**Option 2: Ollama**

1. Install from: https://ollama.ai/

### 🎯 Advanced Controls (Text-to-Image v1.7)2. Pull a model: `ollama pull llama3`

- **Prompt Length**: 6 options (very_short to very_long)3. Server runs automatically (default: http://localhost:11434)

- **Genre/Style**: 22 genres (surreal, cinematic, horror, cyberpunk, noir, etc.)

- **Subject Framing**: 14 shot types (close-up, wide shot, cowboy shot, etc.)## Node Inputs

- **Subject Pose**: 17 poses (standing, action, contrapposto, etc.)

- **Camera**: Angle, composition (11 options each)### Core Inputs

- **Lighting**: Source, quality, time of day, weather (8-13 options each)- **basic_prompt**: Your simple video idea

- **Color**: Art style, color mood (18, 10 options)- **preset**: Style preset (custom/cinematic/surreal/action/stylized/noir/random)

- **Detail Level**: 6 options from simplified to intricate- **expansion_tier**: Detail level (auto/basic/enhanced/advanced/cinematic)

- **mode**: text-to-video or image-to-video

All controls support: `auto` (intelligent defaults), `random` (variety), `none` (disable), or specific values.

### LLM Configuration

### 🔧 Special Syntax Support- **llm_backend**: lm_studio or ollama

- **model_name**: Model identifier (e.g., "llama3")

**Emphasis (Weight Control)**- **api_endpoint**: API URL (default LM Studio: http://localhost:1234/v1)

```- **temperature**: 0.1-2.0 (lower = focused, higher = creative)

(keyword:1.5)    # Increase importance

(keyword:0.5)    # Decrease importance### Keywords

```- **positive_keywords**: Comma-separated must-include terms (LoRA triggers, etc.)

- **negative_keywords**: Comma-separated terms to avoid

**Alternation (Random Selection)**

```### Output Options

{cat|dog|rabbit}           # Picks one randomly- **num_variations**: Generate 1-3 variations

{red|blue|green} dress     # Random color- **save_to_file**: Save prompts to disk

```- **filename_base**: Base name for saved files



**Combined**## Node Outputs

{elegant|casual} woman with (detailed face:1.5) and {blonde|red} (hair:1.2)1. positive_prompt_1: First enhanced prompt


3. **positive_prompt_3**: Third variation (if requested)

---4. **negative_prompt**: Auto-generated negative prompt

5. **breakdown**: Detailed analysis of expansion

## 📦 Installation6. **status**: Success/error messages and file save location



### Method 1: Git Clone (Recommended)## Expansion Tiers



```bash### Basic

cd ComfyUI/custom_nodes/**Formula**: Subject + Scene + Motion

git clone https://github.com/EricRollei/video_prompter.git- Simple expansion with essential details

```- Clear subject, setting, and action

- ~50-100 words

### Method 2: Manual Install

### Enhanced

1. Download this repository as ZIP**Formula**: Subject + Scene + Motion + Basic Aesthetics

2. Extract to `ComfyUI/custom_nodes/video_prompter/`- Detailed descriptions with characteristics

3. Restart ComfyUI- Basic shot size and lighting

- ~100-200 words

### Dependencies

### Advanced

Install Python requirements:**Formula**: Full Details + Complete Aesthetics + Camera Work

- Professional cinematography terms

```bash- Specific lighting, composition, lens choices

cd ComfyUI/custom_nodes/video_prompter/- Camera angles and movements

pip install -r requirements.txt- ~200-350 words

Cinematic

**Requirements:**Formula: Masterful Description + All Professional Elements

Python 3.8+- Director-level shot description
PyTorch (usually included with ComfyUI)- Complete lighting setup with technical details
NumPy (usually included with ComfyUI)- Precise camera movements and choreography
Pillow/PIL (usually included with ComfyUI)- Color grading and atmospheric elements
requests (for LLM API calls)- ~350-500+ words

---## Presets Explained

⚙️ Setup### Cinematic

Professional film quality with emphasis on lighting, composition, and camera movement.

LLM Backend Setup (Required)- Edge lighting, soft lighting, warm colors

Smooth camera movements

Choose one of these options:- Professional framing

Option 1: LM Studio (Recommended)### Surreal

Dreamlike, otherworldly aesthetics with unusual elements.

Download: https://lmstudio.ai/- Unnatural lighting and mixed sources
Install and launch LM Studio- Floating, ethereal movements
Download a model (search for "llama" or "mistral", recommended: Llama 3.2 8B)- Saturated or desaturated colors
Start the server: Click "↔" (Server) tab → "Start Server"
- Default: http://localhost:1234/v1### Action
Note the model name (e.g., "llama3")High-energy, dynamic sequences with rapid motion.

Fast camera movement, tracking shots

See docs/LM_STUDIO_SETUP.md for details.- High contrast, dramatic lighting

Explosive, kinetic action

Option 2: Ollama

Stylized

Install: https://ollama.ai/Artistic visual style over realism.
Pull a model: ollama pull llama3- Strong visual identity
Server runs automatically at http://localhost:11434/v1- Consistent color palette

Graphic, illustrative elements

Noir

🚀 Quick StartDark, moody, high-contrast aesthetic.

High contrast, hard lighting

Add Node: Right-click → Add Node → Eric Prompt Enhancers- Deep shadows and chiaroscuro
Enter Prompt: "a woman in a garden"- Desaturated or black and white
Configure LLM: Backend (lm_studio), Model Name (llama3), API Endpoint- Low angles and dutch angles
Set Options: Platform, style, settings
Connect Output: Link to your generation node### Random
Generate: Run workflowRandomly selects complementary elements from the Wan 2.2 guide for creative exploration.

Example Output## Usage Examples


Input: "a woman in a garden"```

Platform: fluxInput: "A cat playing piano"

Genre: cinematicPreset: Cinematic

Length: mediumTier: Auto (detects "Enhanced")



Output: "Cinematic medium shot portrait of an elegant woman in a lush garden Output: "A fluffy orange tabby cat sits at a polished black grand piano in a warmly lit living room. The cat's paws press down on the ivory keys with surprising dexterity. Soft afternoon light streams through a nearby window, creating edge lighting on the cat's fur. Medium close-up shot, eye-level angle. The camera slowly pushes in as the cat continues to play. Warm color palette, soft lighting, clean single shot."

bathed in warm golden hour sunlight, captured from a low angle, graceful pose, ```

detailed botanical environment with soft bokeh, professional photography quality, 

(detailed face:1.4), flowing dress, film-like aesthetics"### Example 2: Image-to-Video with Keywords

Input: "The woman turns and walks toward the camera"

---Mode: Image-to-Video

Preset: Noir

📚 DocumentationPositive Keywords: myLoRA_trigger, detailed_face

Tier: Advanced

Comprehensive guides in docs/:

Output: "The woman slowly turns her head, edge lighting creating dramatic shadows across her face, myLoRA_trigger, detailed_face. She begins walking directly toward the camera with deliberate, purposeful steps. Her silhouette is backlit by a single streetlight, creating a strong rim light effect. The camera remains static as she approaches, her features gradually emerging from shadow. High contrast lighting, desaturated colors, low angle shot. The scene has a tense, mysterious atmosphere."

Getting Started```

QUICKSTART.md - Fast introduction
LM_STUDIO_SETUP.md - LM Studio setup### Example 3: Multiple Variations for Action
CONFIGURATION.md - Settings explained```

Input: "Superhero landing from the sky"

Node GuidesPreset: Action

TXT2IMG_GUIDE.md - Text-to-Image (20+ pages)Variations: 3
IMG2IMG_GUIDE.md - Image-to-ImageTier: Cinematic
WILDCARD_GUIDE.md - Random wildcards

Results in 3 different takes:

Reference- Variation 1: Emphasizes impact crater and dust

QUICK_REFERENCE.md - All nodes- Variation 2: Focuses on dynamic camera arc around landing
NODE_COMPARISON.md - Which node to use- Variation 3: Features slow-motion with particle effects


### Updates

- [CHANGELOG.md](CHANGELOG.md) - Version history## Workflow Integration

- [UPDATE_V17_ENHANCED_CONTROLS.md](docs/UPDATE_V17_ENHANCED_CONTROLS.md) - v1.7

### Basic Workflow

---```

[AI Video Prompt Expander]

## 🔄 Version History    ↓ (positive_prompt_1)

[Video Generation Node]

### v1.7.0 (October 2025) - Enhanced Controls    ↓

- ✅ Fixed reference image usage[Output]

- ✨ Prompt length control (6 options)```

- ✨ Genre/style control (22 genres)

- ✨ Subject framing (14 types)### Advanced Workflow with Variations

- ✨ Subject pose (17 options)```

[AI Video Prompt Expander] (num_variations: 3)

### v1.6.1 (October 2025) - Bug Fixes    ↓ (positive_prompt_1) → [Video Gen 1]

- 🐛 Fixed settings leaking into output    ↓ (positive_prompt_2) → [Video Gen 2]  

- 🐛 Fixed emphasis syntax `(keyword:1.5)`    ↓ (positive_prompt_3) → [Video Gen 3]

- ✨ Added alternation `{a|b|c}`    ↓ (negative_prompt) → [All Video Gens]

v1.6.0 (October 2025) - Text-to-Image

✨ New Text-to-Image node## File Output
🎨 8 platform support
📸 Reference imagesWhen save_to_file is enabled, prompts are saved to:


See [CHANGELOG.md](CHANGELOG.md) for complete history.ComfyUI/output/video_prompts/[filename_base]_[timestamp].txt

File includes:

🛠️ Troubleshooting- Enhanced positive prompt

Negative prompt

LLM Connection- Detailed breakdown

Check server is running (LM Studio "Running" status)- Metadata (preset, tier, model, temperature)
Verify endpoint URL matches- Original input
Model name exact match (case-sensitive)

Troubleshooting

Output Quality

Adjust prompt_length for target size### "Cannot connect to LM Studio/Ollama"
Verify genre_style matches mood- Ensure LM Studio or Ollama is running
Update to v1.6.1+ for bug fixes- Check the API endpoint URL
Verify the model is loaded

Node Not Appearing

Restart ComfyUI after installation### "Request timed out"
Check "Eric Prompt Enhancers" category- LLM might be slow on complex prompts
Run pip install -r requirements.txt- Try lowering the tier
Reduce temperature

---- Use a smaller/faster model

📄 License### "Missing keywords in output"

Keywords are automatically appended if missing

Dual License:- Check the positive_keywords field spelling

Verify keywords are comma-separated

Non-Commercial Use

Creative Commons Attribution-NonCommercial 4.0 International### Empty outputs

Check LLM backend connection

Commercial Use- Review status output for error messages

Separate license required. Contact:- Try with a simpler input prompt first

Email: eric@historic.camera or eric@rollei.us
GitHub: @EricRollei## Tips for Best Results

See LICENSE for full terms.1. Start Simple: Begin with basic concepts and let the node expand them

Use Auto Tier: Let the node detect appropriate complexity

---3. Experiment with Presets: Different presets dramatically change the output

Adjust Temperature: Lower (0.3-0.5) for consistency, higher (0.8-1.2) for creativity

🙏 Credits5. Combine Keywords: Use positive_keywords for LoRA triggers and style terms

Save Good Prompts: Enable save_to_file to build a library

Author7. Generate Variations: Use multiple variations to explore different interpretations

Eric Hiss (@EricRollei)

Wan 2.2 Guide Reference

Dependencies

ComfyUI - GPL-3.0This node implements the comprehensive Wan 2.2 prompting framework including:
PyTorch - BSD-style
NumPy - BSD- Aesthetic Control: Lighting sources/types, time of day, shot sizes, composition, lenses, camera angles
Pillow - HPND- Dynamic Control: Motion types, character emotions, camera movements
requests - Apache 2.0- Stylization: Visual styles (3D, 2D, watercolor, etc.), visual effects

LLM BackendsFor more details, see the original guide: https://wan.video/

LM Studio - Proprietary
Ollama - MIT## Credits

Platforms- Based on the Wan 2.2 Video Generation Prompting Guide

Flux by Black Forest Labs- Developed for ComfyUI integration
SDXL by Stability AI- Supports LM Studio and Ollama backends
Pony Diffusion by Astralite
Illustrious XL by OnomaAI Research## Version
Chroma/Meissonic by Tencent AI Lab
Qwen by Alibaba Cloudv1.0 - Initial release with full Wan 2.2 framework integration
Wan by Wuhan AI Institute

License

MIT License - Free to use and modify

📞 Support

GitHub Issues: Report bugs here
Email: eric@historic.camera, eric@rollei.usQuestions or Issues? Check the status output for detailed error messages, or review the saved prompt files for debugging information.
Documentation: Check docs/ folder

Made with ❤️ by Eric Hiss | Star ⭐ if you find this useful!

README.md Unescape Escape

Eric's Prompt Enhancers for ComfyUI# Eric's Prompt Enhancers for ComfyUI