You can generate custom AI images directly from your Files page. Loophole creates the image based on your prompt and saves it to your account as a PNG file.
For a complete gallery of use cases - including timelines, flowcharts, annotated documentation, and more - see: Generate Image: A Gallery of Ideas
How to generate an AI image
Go to your Files page
Navigate to the Files section in your account dashboard.Open the image generation tool
Click the dropdown arrow next to "Upload Files" and select "Generate Image."Choose your aspect ratio
Select the dimensions that fit your needs:Square - equal width and height
Landscape - wider than tall
Portrait - taller than wide
Enter your prompt
Describe the image you want to create. Be specific about what you want to see.Add a reference image (optional)
If you want the AI to use an existing image as visual context, click the reference image selector and choose a file from your Files.
This is entirely optional - you can generate images with just a prompt.Don't have the image uploaded yet? Just upload it to your Files first (using the normal upload process), then come back to Generate Image.
Note: Using a reference image adds $0.25 to the generation cost ($0.50 total).Click generate
The image is generated automatically - typically takes 10-30 seconds depending on server load.Image saved
The generated image appears in your Files list as a PNG.
What happens after generation?
Thumbnail appears
Shows a preview of your generated image
Visible within a few seconds of completion
Loophole AI analyzes the image
Generates a descriptive title based on content
Creates a detailed description of what's shown
Categorizes the content type
All file features available
Add notes, rename it, associate with a business
View full-size PNG
Download original file
All activity logged
Reuse your prompt
Open your generated image and go to the AI Data tab. Your original prompt is displayed in the Prompt field. Click the reuse icon (↻) in the bottom-right corner of the prompt, and it will open the Generate Image dialog with your prompt already filled in - ready to regenerate or tweak. Then click 'Generate Image' when you are ready.
Tips for writing effective prompts
The quality of your output depends heavily on your prompt. Here's how to get better results:
Be specific, not vague
❌ "A car in a parking lot"
✅ "A red sports car parked alone in an empty parking lot at sunset, with long shadows stretching across the pavement"
Describe the style
Mention if you want it to look like a cartoon, illustration, photograph, sketch, infographic, etc.
Example: "A cartoon-style illustration of..." or "A professional infographic showing..."
Include details about composition
Where should elements be positioned?
What's in the foreground vs. background?
What perspective or angle?
Specify colors when they matter
If color is important to your message, say so explicitly
Example: "A green checkmark next to a red X"
Describe the mood or tone
Is it humorous? Serious? Professional? Playful?
Example: "A lighthearted, satirical cartoon depicting..."
Think about what you'll use it for
An image for a formal complaint should look different than one meant to add humor
Let the purpose guide your style choices
Use a reference image when it helps
If you're illustrating something about a specific place, object, or situation, uploading a reference image can save you a lot of descriptive work. Instead of trying to describe your apartment layout in words, just show it. The AI will incorporate the visual context into your generated image.
If you're not sure how to prompt
If you're working on an active Problem with Loophole AI & Tools, you don't have to figure out the prompt yourself.
Research Console can analyze your situation, recognize when a visual would strengthen your communication, and draft a detailed image generation prompt for you. You review it, modify it if you want, and approve it before anything is generated.
Research Console is only available during an active Problem - it's not accessible for manual image generation from the Files page. But if you're stuck on what to create or how to describe it, starting a Problem gives you access to AI-powered creative strategy.
For more on how Generate Image fits into your communication strategy, see: When Words Aren't Enough: Using Generate Image
Pricing
$0.25 per image generated from the Files page. An additional $0.25 fee applies when using a reference image ($0.50 total).
Deducted from your available credit balance
You must have sufficient balance to generate - if your balance is below $0.25, the generation won't proceed
No daily or monthly limits on how many images you can generate
Within an active Problem: Image generation is handled by Create Console and uses your Tool Action Credit tokens instead of the $0.25 fee.
Refund policy
No refunds for images you don't like.
If the generated image doesn't match what you expected, you'll need to refine your prompt and generate a new image (which costs another $0.25).
Why? Loophole doesn't receive refunds from Google when an image is successfully generated - so we can't offer them either.
The only exception: If image generation completely fails - meaning all automatic retries are exhausted and the system is unable to produce any image - your $0.25 will be automatically refunded.
Tip: Take time to write a detailed, specific prompt before generating. If you're not happy with the result, open the image, go to the AI Data tab, and click the reuse icon (↻) to reload your prompt into the generator - then refine it and try again.
Keep in mind, each generation is a new charge ($0.25, or $0.50 with a reference image) - reusing a prompt doesn't waive the fee.
If you're unsure where to start, consider working within an active Problem where Research Console can help craft the prompt for you.
What you cannot generate
Generate Image is a communication tool - not an evidence fabrication tool.
Prohibited content includes:
Fake photos of damage, incidents, or conditions
Images that misrepresent actual facts or conditions
Fraudulent "evidence" of any kind
Content designed to deceive anyone about what actually happened
Using reference images to make fabricated content appear more authentic or realistic
Uploading a real photo and prompting the AI to add fake damage, conditions, or evidence that doesn't exist
Any use of reference images intended to deceive someone about what actually happened
Illegal content, violence, or explicit material
Content that violates Loophole's Terms and Conditions
If you attempt to generate prohibited content:
Your account will be immediately flagged
Your access will be disabled while an admin reviews your activity
If misuse is confirmed, your account will be permanently locked
You will not be able to create a new Loophole account
Generate Image helps you illustrate real situations creatively. It never helps you fabricate false ones. This applies equally to reference images - you cannot use a real photo of your property to generate fake damage, conditions, or "evidence" of things that didn't happen.
Storage
Generated images count toward your storage plan:
Saved as PNG images
File size varies based on complexity and aspect ratio
Counts toward your storage limit
Account requirements
Active accounts only. Image generation requires an Active account status.
If you're on the Waitlist and need to generate an image, you can activate your account by completing the application process and adding funds.
Limitations
Generation time varies
Typically 10-30 seconds
May take longer during high server load on OpenAI or Google
AI interpretation
The AI interprets your prompt - results may not match your vision exactly
More specific prompts produce more predictable results
Retry behavior
If generation fails, the system will automatically retry
After all retries are exhausted, you'll be notified and refunded
Style consistency
Different prompts may produce different styles
If you need a consistent look across multiple images, use the same model and similar prompt structure