How to Use NPC Forge

NPC Forge is an AI-powered character generator that creates detailed NPCs through a step-by-step wizard interface. This comprehensive guide covers character creation, library management, interactive chat, and editing features.

Getting Started

First Steps

1Visit NPC Forge and you'll see a welcome popup for new users
2Click "Get Started" to begin the character creation wizard
3The application is completely free to use with built-in AI capabilities
4All character data is stored locally in your browser for privacy

Quick Character Creation

Click "Generate Random Character" from any step for instant results
This creates a character with default settings and takes you to the final step
Perfect for testing the system or when you need inspiration

Character Creation Wizard

Step 1: Concept

Choose Your Foundation

Concept Step Interface

Select a Genre: Fantasy, Sci-Fi, Historical, or Contemporary
Pick a Sub-Genre: 16 specialized templates like High Fantasy, Cyberpunk, Medieval, Mystery & Thriller
Write Your Description: Be specific about appearance, personality, and background

Example Description: "A scarred elven ranger who protects a sacred forest, harboring a secret connection to ancient magic that causes plants to grow in her footsteps."

Tips for Better Results:

Include visual details for better portraits
Mention personality traits or quirks
Add context about role or background
2-3 sentences work better than paragraphs

Step 2: Options

Customize Character Details

Options Step Interface

Basic Traits:

Gender (Male, Female, Nonbinary, Unknown)
Age Group (Child, Teen, Adult, Elder)
Moral Alignment (Good, Neutral, Evil)
Relationship to Player (Ally, Enemy, Neutral, Mentor, Rival, Betrayer)

Advanced Options (expandable section):

Physical Traits: Height, build, distinctive features
Background: Social class, homeland/origin
Occupation: Searchable dropdown with genre-specific options
Personality Traits: Multi-select system (unlimited selection)

Content Generation Options:

Include Quests: 3-7 quests with customizable types and rewards
Include Dialogue: 5-15 character-appropriate dialogue lines
Include Items: 3-10 items with rarity and category settings
Include Portrait: Toggle AI-generated character artwork

Using the Controls:

Randomize: Generates random traits while preserving your description
Clear Options: Resets all fields except description and portrait settings
Progress bar at top allows jumping between steps

Step 3: Model Selection

Choose AI Quality and Usage

Model Selection Interface

Text Models (for character content):

Standard (gpt-4o-mini): 50 generations/month, good quality
Enhanced (gpt-4.1-mini): 30 generations/month, better detail
Premium (gpt-4o): 10 generations/month, highest quality

Image Models (for portraits):

Standard (dall-e-2): 10 generations/month, basic portraits
Enhanced (dall-e-3): 5 generations/month, improved quality
Premium (gpt-image-1): 3 generations/month, best quality + editing support

Portrait Customization Options:

Art Style: Realistic, Fantasy Art, Anime, Comic Book, Oil Painting
Expression: Neutral, Happy, Serious, Angry, Determined, Mysterious
Framing: Portrait, Bust, Full Body, Action Pose
Background: Plain, Gradient, Themed, Environmental, Abstract

Model Selection Tips:

Start with Standard models for experimentation
Use Enhanced/Premium for important characters
Consider mixing tiers (Standard text + Enhanced image)
Monitor usage indicators to track monthly limits

Step 4: Generate

Create Your Character

Generation Results Interface

Click "Generate Character" to start AI processing
Generation takes 30-120 seconds depending on options and model selection
Loading message appears after 3 seconds for longer generations
Results display in tabbed interface: Profile, Quests, Dialogue, Items

After Generation:

Save to Library: Stores character locally for editing and chat
Download JSON: Exports character data for backup
Download Portrait: Saves portrait image (if generated)
New Character: Starts wizard over with fresh options

Character Library

Saving and Organizing Characters

Character Storage:

All characters saved locally in your browser using IndexedDB
Includes character data, portraits, and later chat conversations
No data is sent to external servers
Characters persist across browser sessions

Library Interface:

Character cards show portrait, name, and trait tags
Action buttons on each card: Chat, Edit, Download JSON, Download Portrait, Delete
Click character cards to open detailed modal view
Search bar at top for finding specific characters

Enhanced Filtering and Search

Automatic Trait Discovery:

Filter dropdowns automatically created from your character data
Organized into categories: Basic Info, Physical Traits, Background, Personality
New traits appear in filters as you save more characters

Smart Search Functionality:

General text search across all character data
Trait-specific syntax: personality: brave, occupation: knight, genre: fantasy
Combine multiple filters and search terms
Results update in real-time as you type

Using Filters Effectively:

Use dropdown filters to narrow by trait categories
Combine with search for precise results like "fantasy characters who are mentors"
Filter panels are collapsible for better organization
Clear all filters with reset button

Character Editing

Accessing Edit Mode:

Click "Edit" button on character cards
Edit page provides comprehensive character modification tools
Model selectors at top allow choosing different AI models for regeneration

Available Edits:

Basic Information: Name, description, core traits
Character Content: Modify quests, dialogue, and items
Add/Remove Elements: Add new quests, dialogue lines, or items
Individual Regeneration: Regenerate specific character elements
Portrait Management: Upload, regenerate, or edit portraits
Trait Management: Add, edit, or regenerate character traits

Regeneration Features:

Individual attributes (name, appearance, personality, backstory)
Quest components (title, description, reward separately)
Dialogue lines and item descriptions
Portrait regeneration with different models
Each regeneration counts against selected model's monthly limit

Import and Export

Exporting Characters:

Click "Download JSON" on character cards
Files contain complete character data including traits and portrait information
Use for backup, sharing, or moving between devices

Importing Characters:

Use import option in library interface
Select JSON files from your device
Characters added to library with all data intact
Trait data automatically integrated into filtering system

Interactive Chat System

Starting Conversations

Multiple Ways to Begin:

Click "Chat" button on character cards in library
Use "Start Chat" button in character detail modal
Navigate directly to /chat/[characterId] URL
Chat links available in enhanced sidebar navigation

Chat Interface Features:

Character portrait and name in compact header
Model selection dropdown (Standard/Enhanced/Premium)
Message input with 1000 character limit and counter
Scrollable conversation history
Clear chat and retry message options

How Chat Works

Character Consistency:

AI maintains character personality using all trait data, backstory, and relationships
Recent conversation context (last 15 messages) provided for continuity
Responses reflect character occupation, background, and personality traits
System prompts ensure character stays in character throughout conversation

Dynamic Response System:

Simple greetings: Brief responses (1-2 sentences)
Casual questions: Moderate responses (2-4 sentences)
Detailed requests: Fuller responses (1-2 paragraphs)
Maximum 3 paragraphs to prevent response cutoff

Model Quality Differences:

Standard: Basic personality consistency, good for casual conversation
Enhanced: Better character voice and depth, improved context understanding
Premium: Sophisticated personality depth, excellent consistency, nuanced responses

Conversation Management

Persistent Storage:

All conversations saved locally per character
History survives browser restarts and device changes
Maximum 100 messages per session (automatically trimmed)
Each character maintains separate conversation history

Usage Integration:

Each chat response counts as one generation against text model monthly limit
Can switch between model tiers mid-conversation
Usage warnings appear when approaching monthly limits
Model selection affects response quality

Chat Best Practices:

Start with simple greetings to establish character voice
Ask about character background and motivations
Reference character traits in your questions for consistency
Use higher-tier models for important story moments
Switch models based on conversation importance

Portrait Editing

Accessing Portrait Editing

Requirements:

Character must have an existing portrait
Only gpt-image-1 (Premium image model) supports full editing
dall-e-2 and dall-e-3 do not support editing at all

Edit Interface:

Available in character edit pages for compatible characters
Text area for describing desired changes
Character counter shows remaining prompt length (varies by model)
Model compatibility warnings for unsupported models

Using Portrait Editing

Writing Effective Edit Prompts:

Be specific about changes: "change hair color to blonde", "add glasses"
Focus on one major change per edit: avoid complex multi-part edits
Use clear language: "make them smile" rather than "happier expression"
Examples of successful edits:
Color changes: "change shirt to red", "make eyes green"
Accessories: "add a hat", "remove glasses", "add armor"
Expressions: "make them smile", "give serious expression"

Edit Process:

Enter clear description of desired changes
Click "Edit Portrait" to start processing
Editing takes 30-90 seconds
Results appear when processing completes
Original portrait preserved until you save changes

Usage and Limits:

Portrait edits count against image model monthly limits
gpt-image-1 supports up to 32,000 character prompts
Usage warnings appear when approaching limits
Failed edits don't count against usage

Advanced Trait Management

Additional Traits System

Trait Display:

All additional character traits shown in organized layout
Consistent Title Case formatting throughout interface
Editable names and values for custom traits
Empty state displayed when no additional traits exist

Adding New Traits:

"Add Generated Trait": AI creates contextually appropriate traits
Custom Traits: Manually add trait names and values
Both options create properly formatted, editable entries
New traits integrate with existing display and filtering

Managing Existing Traits:

Edit trait names and values in-place
Individual regenerate buttons for each trait
Delete unwanted traits
Multiple trait operations can be performed in sequence

AI Trait Generation

How It Works:

AI analyzes character personality, background, and occupation
Generates traits that enhance roleplay and character depth
Maintains consistency with existing character elements
Produces appropriately formatted trait entries

Usage Considerations:

Trait generation counts against text model monthly limits
Generated traits match character context and genre
Can regenerate individual traits if results don't fit
Traits integrate with library filtering system

Tips for Best Results

Character Creation Strategy

Start Simple: Create basic character, then enhance with editing features
Be Specific: Detailed descriptions produce more tailored characters
Use All Content Types: Quests, dialogue, and items add roleplay depth
Plan Model Usage: Use Standard for testing, Enhanced/Premium for important characters

Library Organization

Consistent Naming: Use descriptive character names for easy searching
Strategic Traits: Apply relevant traits for better filtering
Regular Backups: Export important characters as JSON files
Trait Management: Use AI generation to add depth to existing characters

Chat Development

Start Conversations Early: Begin chatting immediately after character creation
Explore Backstory: Ask about character history and motivations
Reference Traits: Mention character occupation and background in conversations
Model Strategy: Use appropriate tiers based on conversation importance

Portrait Workflow

Generate First: Start with AI portrait generation
Edit Incrementally: Make small adjustments rather than major changes
Plan Edits: Consider what visual elements would enhance your character
Use Premium Model: gpt-image-1 provides best editing results

Troubleshooting Common Issues

Character Generation

Failed Generation: Check internet connection and try again
Unexpected Results: Try more specific descriptions or different trait combinations
Slow Generation: Portrait generation can take up to 2 minutes

Chat Problems

Chat Not Loading: Verify character exists in library and browser supports IndexedDB
Messages Not Sending: Check usage limits and network connection
Inconsistent Responses: Use higher-tier models or add more personality details

Portrait Issues

Edit Button Disabled: Ensure character has existing portrait and compatible model selected
Editing Failed: Check usage limits, try simpler prompts, verify network connection
Poor Results: Use gpt-image-1 model and more specific edit descriptions

Library Issues

Characters Not Saving: Check browser storage permissions and available space
Search Not Working: Clear browser cache and refresh page
Missing Characters: Verify IndexedDB isn't corrupted