Full Tour

🎨 Introduction & Overview

🤖 What is the AI Image Generation Tool

The AI Image Generation Tool is a powerful interface that enables you to create and transform images using state-of-the-art artificial intelligence models. Whether you're enhancing fabric presentations, creating marketing materials, or exploring creative possibilities, this tool provides everything you need to generate professional-quality visual content.

Key capabilities include:

🖼️ Image-to-Image Transformation: Use an existing image as a base and transform it with AI
🎬 Video Generation: Convert static images into dynamic video content
🎯 Inpainting: Edit specific regions of images with precision control

🪙 Token System Overview

The AI Image Generation Tool operates on a token-based system, allowing you to control your usage and costs transparently.

How tokens work:

Each generation consumes a specific number of tokens based on the model and settings
Your token balance is displayed prominently in the interface
Tokens are deducted when a generation starts
Real-time balance updates keep you informed of your remaining credits

Token costs vary by model and feature:

Image generation models have per-image token costs
Video generation costs are calculated per second of video
Batch size multiplies the base token cost
All costs are displayed before you generate

You can check your current token balance at any time in the Settings Panel, and the system will alert you when your balance is insufficient for a generation.

🚀 Getting Started - Interface Layout

The AI Image Generation Tool interface is organized into four main sections, each designed for specific tasks in your creative workflow.

📍 Main Interface Sections

🎨 Left/Center: Image Editor Canvas The central workspace where your images appear and can be edited. This canvas supports multiple viewing and editing modes, allowing you to work with images in the way that best suits your workflow.

⚙️ Right: Settings Panel Contains either the Prompt Panel for image/video generation or the Inpainting Panel for precision editing. This is where you configure your generation parameters, write prompts, and control AI behavior.

📚 Bottom: History Gallery A scrollable gallery showing all your previous generations, both images and videos. Click any completed image to load it back into the canvas for further editing or use.

🖼️ Image Editor Canvas

The canvas is your primary workspace for viewing, editing, and preparing images for AI generation.

📥 Loading an Image

There are three convenient ways to load images into the canvas:

📂 Option 1: Click to Browse

Click anywhere on the empty canvas area
The Image Browser modal will open
Select an image from your library

🖱️ Option 2: Drag and Drop

Drag an image file from your computer
Drop it directly onto the canvas
Supports all common image formats (JPG, PNG, WebP, etc.)

🔍 Option 3: Browse Button

Click the "Browse Images" button (visible in edit mode header)
Access images from multiple sources: uploads, AI outputs, materials, or Cutout Studio

👁️ Viewing Modes

The canvas supports two distinct modes to separate viewing from editing:

Static View Mode (Default)

Image displays without editing capabilities
Clean, distraction-free viewing
Click the "Edit" button to enable editing features
Perfect for reviewing generations or using images as-is

✏️ Edit Mode

Fabric.js canvas overlays the image
Enables adding objects, drawing, and manipulation
Click the X button to exit edit mode and return to static view
Changes made in edit mode are captured in the final generation

🎛️ Canvas Controls

🧹 Clear Canvas Button Located in the top right corner (X icon), this button instantly clears the canvas and resets it to empty state, ready for a new image.

📐 Aspect Ratios

The canvas dynamically adjusts to your selected aspect ratio from the Settings Panel. Visual feedback shows you exactly how your image will be cropped or fitted to match the chosen ratio. This ensures your generated output matches your intended dimensions perfectly.

⚙️ Settings Panel (Prompt Panel)

The Settings Panel is your control center for configuring AI generation parameters. It adapts based on the selected model, showing only relevant options.

✍️ Prompt Input

The large text area at the top is where you describe the transformation or image you want to create.

📏 Capacity: 2500 character limit
⚡ Generation: Press Enter to start generation
↵ New Lines: Use Shift+Enter to add line breaks within your prompt
✨ Best Practices: Be specific, descriptive, and include details about style, mood, lighting, and composition

🎲 Random Prompt Generator

Need inspiration or want to explore what's possible? Click the dice icon button next to the prompt label to generate creative sample prompts. The generator provides different prompts depending on whether you're using image or video models, giving you relevant starting points for your generation.

🤖 Model Selection

Choose from multiple AI models, each optimized for different tasks:

**Image **

Generates images from prompts without requiring an input image
Supports up to 3 additional reference images for style guidance
Token cost displayed next to model name

**Video **

Creates video content from a static image plus prompt
Cost calculated per second of video duration
Currently supports 16:9 aspect ratio only

💰 Model-Specific Features Each model displays its token cost clearly, helping you make informed decisions about which model to use for your project.

📐 Aspect Ratio Selection

Choose from six standard aspect ratios to match your intended use case:

🖥️ 16:9 - Desktop, widescreen presentations
⬛ 1:1 - Square, social media posts
📱 9:16 - Mobile, vertical video
📺 4:3 - Classic displays
📄 3:4 - Portrait orientation
🎬 21:9 - Ultrawide, cinematic

Note that some models have restrictions (e.g., Video only supports 16:9).

🖼️ Additional Reference Images

Available for Image model only

Add up to 3 additional reference images to guide the AI's style, composition, or aesthetic:

📊 Current Count Display: Shows "0/3", "1/3", etc.
🎛️ Grid Layout: Reference images display in an organized grid
➕ Add Images: Click the + button to open Image Browser
❌ Remove Images: Click the X on any image to remove it
🖱️ Drag and Drop: Supported for quick addition

Reference images help maintain consistency across generations and provide visual examples of your desired output style.

🔢 Batch Size

Available for Image models only (not Video)

Generate multiple variations in a single generation:

1️⃣ Options: 1, 2, 4, 8, or 16 images
💰 Token Multiplication: Cost multiplies by the number of images
💡 Use Case: Generate 1 image to test prompts, then increase batch size when satisfied

⏱️ Video Duration

Available for Video model only

Select the length of your generated video:

🎬 Options: 4s, 6s, or 8s
🪙 Token Cost: 55 tokens per second (e.g., 4s = 220 tokens, 8s = 440 tokens)
🎯 Strategy: Start with shorter durations for testing, increase when satisfied with results

🚀 Generate Button

The main action button that starts your AI generation:

🏷️ Label: Shows "Generate" or "Generate Video" based on selected model
🔒 State Changes: Disabled when requirements aren't met
✅ Required Conditions:
- Image loaded (when required by model)
- Prompt entered
- Sufficient tokens available
- No generation currently in progress

⚠️ Insufficient Tokens Warning

When your token balance is too low for a generation, the Generate button is replaced with a warning message and inline token purchase options. This ensures you're never surprised by an inability to generate and provides immediate access to purchase more tokens.

⚡ Generation Process

🎬 Starting Generation

When you click the Generate button:

📢 A toast notification appears: "Starting X image generation(s)..."
💸 Tokens are immediately deducted from your balance
🔄 The generation job is created and begins processing
✨ The UI remains responsive, allowing you to start additional generations

📊 Progress Indicators

The tool uses different progress indication methods depending on the AI model:

🌊 Websocket-Based Models (Older streaming models)

Progressive image streaming shows results forming in real-time
Blur effect gradually decreases as generation progresses
Live updates provide immediate visual feedback

📡 Polling-Based Models (Image, Video)

Status updates check progress every 5 seconds
Shows "Processing..." or "Pending..." status
Final result appears when complete

💰 Token Deduction

Your token balance updates in real-time when generation starts. The exact cost is deducted immediately, and your new balance is displayed in the Settings Panel.

✅ Completion

When generation finishes:

🖼️ The final image appears in the canvas (replacing any previous content)
📚 The History Gallery automatically updates with the new generation
🎉 A success toast may appear (model-dependent)
💾 The image is immediately available for further editing or download

📚 History Gallery

📐 Layout & Overview

The History Gallery is your central hub for managing all AI-generated content:

📍 Location: Scrollable grid at the bottom of the interface
🎨 Content: Combined view of images and videos
♾️ Infinite Scroll: Automatically loads more content as you scroll down
🔢 Item Count: Displayed in the gallery header

🖼️ Image Thumbnails

✅ Completed Images

Display the full generated result
Show selection ring (cyan) when selected
Click to load into canvas for further editing

⏳ In-Progress Images

Display loading spinner with current status
Status text shows "Generating..." or "Pending..."
Update automatically as generation progresses

🎬 Video Thumbnails

Video items are visually distinct from images:

🖼️ Thumbnail: Shows the source image used for video generation
🎞️ Clapperboard Icon: Overlay indicates video content
🔗 Interaction: Clicking opens the video in a new tab
🚫 Selection: Videos cannot be selected or loaded into canvas

📊 Job Status Indicators

Each item clearly shows its current state:

✅ Completed: Full image visible, ready for use
⏳ Processing: Loading spinner + "Generating..." text
⏸️ Pending: Loading spinner + "Pending..." text
❌ Failed: Error indicator showing generation failure

🎯 Interactions

🖱️ Selecting Images Click any completed image to select it. The selected image loads into the canvas and displays a cyan selection ring.

👁️ Viewing Videos Click any video thumbnail to open it in a new browser tab for viewing.

🗑️ Deleting Items Hover over a completed image to reveal the trash icon. Click to delete, confirm in the dialog.

📋 Copying Prompts Hover over any completed item to see a tooltip showing the original prompt. Click the copy icon in the tooltip to copy the prompt to your clipboard and automatically populate the Prompt Panel.

🔄 Loading More History

The gallery uses infinite scroll for efficient loading:

⬇️ Scroll to the bottom of the gallery
🔄 System automatically fetches the next page
⏳ "Loading..." spinner appears during fetch
🏁 "No more to show" message appears when all items are loaded

🔓 Accessing the Browser

Open the Image Browser through any of these methods:

🖱️ Click the empty canvas area
🔍 Click the "Browse Images" button in edit mode header
➕ Click the + button when adding additional reference images

📑 Media Tabs

The browser organizes your images into four categories:

📤 Uploads Your uploaded images - any images you've manually uploaded to the system.

✂️ Cutout Studio Background-removed images from the Cutout Studio tool, perfect for compositing.

🤖 AI Output Previously generated AI images, making it easy to iterate on past generations.

🎨 Materials Fabric and material images from your material library.

📤 Upload Feature

Upload new images directly from the browser:

🖱️ Drag and Drop: Drag image files onto the upload area
📁 Click to Select: Click the upload area to open file picker
📄 Supported Formats: All common image formats accepted

✅ Selecting Images

🔍 Browse through the available images in your selected tab
🖱️ Click any image to select it
🔗 The system converts it to a blob URL to prevent CORS issues
❌ Modal closes automatically
✨ Selected image appears in the canvas ready for use

📱 Mobile Behavior

On mobile devices, the Image Browser displays as a full-screen sheet instead of a modal, providing a native app-like experience. Swipe down to dismiss.

💡 Tips & Best Practices

✍️ Writing Effective Prompts

🎯 Be Specific and Descriptive Instead of "nice fabric," try "soft velvet fabric with deep burgundy color and subtle sheen, photographed in natural window light."

🔑 Mention Key Elements

🎨 Style (realistic, artistic, photographic, illustrated)
😊 Mood (elegant, casual, luxurious, modern)
💡 Lighting (soft, dramatic, natural, studio)
🔍 Details (texture, color, patterns, materials)

🎲 Use the Random Prompt Generator Click the dice icon to see example prompts and learn effective prompt structures.

📚 Learn from History Hover over successful generations in the History Gallery to see their prompts. Click to copy prompts that worked well.

💰 Optimizing Token Usage

🎯 Start Small Begin with batch size 1 to test your prompt and settings. Once you're satisfied with the results, increase batch size for variations.

⏱️ Test Video Durations Use 4-second durations when testing video prompts, then increase to 6s or 8s for final versions.

👁️ Monitor Your Balance Keep an eye on your token balance and plan generations accordingly. Purchase tokens in advance during larger projects.

🎨 Canvas Editing Tips

✏️ When to Use Edit Mode Enter edit mode when you want to add context, annotations, or additional elements before generation. The AI will consider everything visible on the canvas.

👁️ When to Use Static View Use static view when you just want to use an image as-is without modifications. This keeps the interface clean and prevents accidental edits.

📐 Canvas Resolution Remember that the canvas is internally captured at 1024×1024 resolution regardless of display size. Consider this when adding fine details.

🖼️ Additional Reference Images

🎨 Maintain Style Consistency Use reference images with similar styles to guide the AI toward your desired aesthetic consistently.

🔢 Image Limit Maximum of 3 additional images. Choose the most representative examples.

🤖 Model Availability Additional reference images only work with the Image model. Other models don't support this feature.

🔧 Troubleshooting

⚠️ Common Issues

❌ "Please select an image and enter a prompt" Ensure the selected model requires an input image and both the image and prompt are provided.

💸 "Insufficient tokens" Your token balance is too low for the selected generation. Purchase more tokens via the purchase section.

🖼️ Image Not Loading Check your browser console for errors. Try re-uploading the image or selecting it again from the Image Browser.

⏳ Generation Stuck Check the History Gallery for status updates. If status remains "Pending" for more than a few minutes, the job may have failed. Try generating again.

🖱️ Canvas Not Responding Try refreshing the page. If problems persist, clear your browser cache and reload.

🚨 Error Messages

💰 "Insufficient credits to generate image" Your token balance has reached zero. Purchase additional tokens to continue generating.

⚠️ "Failed to create job" API error occurred. Check your internet connection and try again. If the problem persists, contact support.

🎨 "Canvas not properly initialized" The canvas failed to load correctly. Refresh the page to reinitialize the editor.

🪙 "Insufficient tokens" Your current token balance is below the cost required for the selected generation configuration.

💳 Token Purchase & Management

💰 Checking Balance

Your token balance is displayed in the Settings Panel and updates in real-time:

💸 After each generation, the cost is deducted immediately
🔄 Balance refreshes automatically
⚠️ Low balance triggers warning messages

🛒 Purchasing Tokens

When your balance is low or insufficient:

🔔 The purchase section appears automatically
📦 Available token packages are displayed with pricing
✅ Select a package and complete the purchase
⚡ Balance updates immediately upon successful purchase

📊 Token Costs Reference

🖼️ Image Generation

💰 Base cost per image varies by model
✖️ Multiplied by batch size (1, 2, 4, 8, or 16)
⚙️ Additional settings may affect cost

🎬 Video Generation

💰 Base cost per second (typically 55 tokens)
⏱️ Multiplied by duration (4s, 6s, or 8s)
📈 Example: 6-second video = 330 tokens

💡 Cost Display All costs are shown before generation, ensuring you always know exactly how many tokens will be used.

🎯 Next Steps

Now that you've completed the full tour, you're ready to start creating with the AI Image Generation Tool. Here are some suggested next steps:

✍️ Practice with Prompts: Experiment with the random prompt generator and modify prompts to see how changes affect results
🤖 Explore Models: Try different AI models to understand their strengths and characteristics
📚 Build a Library: Save successful generations to reference later and build on your best work
💡 Check Example Uses: See the Example Uses page for creative applications and inspiration

Happy creating! 🎨

On this page