User Guide

Everything you need to know about creating characters, generating content, and getting the most out of the platform.

Quick Navigation

Install the App

xchar is a Progressive Web App (PWA), which means you can install it on your device for a native app-like experience with faster load times and easy access from your home screen.

iOS (iPhone/iPad)

To install on iOS using Safari:

Open xchar.ai in Safari
Tap the Share button (square with arrow pointing up)
Scroll down and tap "Add to Home Screen"
Tap "Add" in the top right corner

The app will appear on your home screen just like a native app. You'll get a fullscreen experience without browser UI.

Android

To install on Android using Chrome:

Open xchar.ai in Chrome
Tap the three-dot menu in the top right
Tap "Add to Home screen" or "Install app"
Tap "Install" to confirm

Chrome may also show an install banner at the bottom of the screen - you can tap that for quick installation.

Desktop

To install on desktop (Chrome, Edge, or other Chromium browsers):

Visit xchar.ai
Look for the install icon in the address bar (usually on the right side)
Click it and select "Install"

The desktop app runs in its own window, making it easy to switch to and keeping your browsing separate from your xchar experience.

Character Creation

There are three different ways to create a character, each suited to different needs and levels of customization.

Simple

A simplified creation wizard that streamlines choices for ease of use. Perfect for beginners or when you want to get started quickly without being overwhelmed by options.

Simple mode guides you through the essential choices while handling the more complex settings automatically.

Advanced

The fully fledged creation wizard for complete customisation. Advanced gives you control over every aspect of your character:

Model: Choose the image generation model that fits your style
Body Type: Select physical characteristics and body type
Personality: Define detailed personality traits and speaking style
Face: Upload or generate custom face images for consistent appearance
Voice: Configure voice settings for text-to-speech

🎨 Choosing a Model

The image model you select dramatically affects the style and type of images generated. Different models produce different visual styles - some are more realistic, others more artistic or anime-inspired.

Some models are more SFW-focused while others are designed for NSFW content. Preview images to see what each model produces.

Advanced uses faceswap technology to maintain consistent facial features across all generated images.

Custom

For users who want complete creative freedom over image generation. Custom does not use faceswap technology.

🎨 Choosing a Model

Like Advanced, you can select the image generation model. Different models produce different visual styles and content levels (SFW vs NSFW). This choice dramatically affects all generated images.

Without faceswap, facial consistency may be slightly less precise, but you gain full control over how images are generated.

Custom relies on a Base Image Prompt system:

How the Prompt System Works

Base Image Prompt: This is prepended to ALL image prompts and defines your character's consistent appearance (face, hair, body, etc.)
User Prompt: Appended after the base prompt - use this for things that change between images (scene, position, expression, etc.)

If clothing is present in your base image prompt, be mindful that it may affect results in images where you want different outfits. Consider keeping clothing out of the base prompt if you want variety.

Edit & Recreate

Want to use another character as a starting point? Visit any character's profile, click the ellipsis menu (⋯), and select "Edit & Recreate" to copy all their creation settings.

This is a great way to learn from existing characters or quickly create variations. You can modify any of the copied settings or create a new character based on those settings.

Editing Character

After creating a character, you can fine-tune their settings at any time. Changes you make will apply to all new conversations created with that character.

Character Card

The character card is the core definition of your character. It includes their description, personality, and behavior guidelines that the AI uses when generating responses.

LLM Settings

Configure which AI model powers your character. Different models have different capabilities, response styles, and costs.

Response Settings

Control how your character responds, including response length preferences and formatting options.

Voice

Give your character a voice! Choose from various voice options that match your character's personality. Voice settings enable text-to-speech for character messages.

Conversations

Each conversation can have its own unique settings that override or extend the base character settings. This lets you explore different scenarios with the same character.

Personality

The personality setting affects how your character responds and how suggested messages are generated for you. You can adjust personality per-conversation to explore different dynamics.

Personality changes both the character's response style AND the suggested messages generated for you. A more flirty personality will suggest more flirty responses from you as well.

Memories

Memories help your character remember important details across conversations. Key things to know:

Memories are appended to the character card and use tokens
They add valuable context that persists across messages
Memories are automatically analyzed for relevance and created when appropriate
You can manually delete memories that are no longer relevant

Keep memories focused on important facts. Too many memories can use up your context window and slow down responses.

Images in Conversations

You can generate images directly within your conversations. The AI uses the last few messages as context to create an optimized prompt for the image.

For full control over image generation, including advanced settings and custom prompts, use the dedicated Generate page instead.

User Settings

Access User Settings from the header while in a conversation. These settings let you customise your chat interface:

Context Length: Set your desired context length for conversations
Show Advanced Settings: Toggle visibility of advanced settings in the chat bar (model selector, temperature, etc.)
Hide Message Suggestions: Remove the suggested message options for a cleaner interface
Hide Microphone Button: Remove the voice input button for a cleaner interface

Customise these settings to create your preferred chat experience - whether you want all the controls visible or a minimal, distraction-free interface.

Generating Images

Create stunning AI-generated images with full control over style, quality, and appearance.

Consistent Character

The standard mode for creating images of your character. This mode automatically prepends your character's attributes to your prompt to maintain visual consistency.

Standard characters: Prepends the character's physical attributes and appearance details
Custom characters: Prepends the base appearance prompt you defined during creation

Your prompt describes the scene, pose, or situation - the character's appearance is handled automatically.

Text to Image

Generate images from text descriptions without any character attributes prepended. This mode does not include faceswap - perfect for scenes, objects, or when you want complete AI generation without face consistency.

Generate Face

Create a unique face that can be used during character creation. Generated faces ensure consistent appearance across all images of your character.

Generate a high-quality face first, then use it when creating your character for the best consistency in all generated images.

Advanced Settings

Fine-tune your image generation with these advanced options:

LoRAs (Low-Rank Adaptations)

LoRAs are style models that modify how images are generated. Each LoRA has a default weight that works well for most cases.

Check CivitAI links for suggested trigger words
Adjust LoRA weights to control style intensity
Combine multiple LoRAs for unique styles

Technical Settings

Steps: More steps = more detail, but slower generation
CFG Scale: How closely to follow the prompt (higher = stricter)
Sampler & Scheduler: Algorithms that affect image style
Seed: Set a specific seed to reproduce exact results
Negative Prompt: Describe what you DON'T want in the image

Quality Options

HD Quality: Upscales the image for higher resolution
HD Fix: Corrects common upscaling artifacts
Batch Size: Generate multiple variations at once

Prompt Generation

Two ways to enhance your prompts:

Enhance: Takes your basic prompt and adds professional photography terms and style suggestions
Random Generate: Creates a completely random prompt for inspiration

Generating Videos

Video Basics

Video generation creates short animated clips from your prompts.

Longer, more detailed text prompts work better for Wan video generation. Use the GPT Agent for help creating effective video prompts.

LoRAs for Video

LoRAs are especially important for video generation. They control motion styles, camera movements, and animation quality.

Check the LoRA page links for optimization tips and recommended settings for each LoRA.

Fast vs Full Quality

⚡ Fast Mode (Recommended)

Usually takes 2-3 minutes
Can take up to 5 minutes
Better cost and speed efficiency
Good quality for most use cases

🎬 Full Quality Mode

Takes up to 15 minutes
Higher quality output
Better if LoRA motion isn't behaving as expected
Use when fast mode isn't giving desired results

HD Quality helps with eye quality in videos. Start with Fast Mode and only switch to Full Quality if you're not getting the results you want.

Images & Gallery

Manage your generated images, share them with the community, and use them for your characters.

Submit to Feed

Share your best creations with the community by submitting them to the public feed. Like other users' content to support fellow creators.

Image Actions

For each image in your gallery, you can:

Like: Save images you enjoy
Copy Prompt: Get the full prompt including any prepended enhancements
Download: Save full quality image to your device
Delete: Remove unwanted images

Setting Thumbnails

Set images as thumbnails for your characters. Important distinction:

Public Thumbnails: Must be approved by moderators before being visible
Private Thumbnails: Visible only to you, no approval needed

Character Posts

Creating Posts

Create posts to showcase your character's content. Posts can include images and videos that highlight your character.

Pricing & Earnings

You can set optional pricing on your character posts.

When users purchase your post, you earn 50% of the purchase price.

If you delete a post, users who have already purchased it will still be able to view it. Deletion only removes the post from public visibility.

Earning Credits

Multiple ways to earn credits on xchar by creating quality content.

Pinned Characters

Characters that get pinned receive more visibility, more messages, and therefore more daily rewards from the message share pool.

What Makes a Character Likely to be Pinned?

Two high-quality thumbnails (non-NSFW)
A high-quality video thumbnail (non-NSFW)
Well-written character description and personality

Feed Earnings

Active participation in the feed with high-quality posts can earn you credits. Quality matters - posts that get engagement will earn more.

Reputation

Your reputation affects your earning potential. If someone buys one of your posts and enjoys it, they're likely to buy more. Building a reputation for quality content leads to more sales over time.

Credits System

Understanding the two types of credits on xchar.

Monthly Credits

Monthly credits are included with subscription plans. They refresh each billing cycle and are used first when you perform actions that cost credits.

Permanent Credits

Permanent credits are purchased separately and never expire. They're used after your monthly credits are depleted. Great for heavy users or as a backup.

Ready to Get Started?

Create your first character and start exploring!

Create Character