Skip to main content

Frequently Asked Questions (FAQ)

Common questions and answers about using Cubby Image.

Installation & Setup​

What is Cubby Image?​

Cubby Image is an AI image generation plugin for Adobe Photoshop that brings image generation, inpainting, and upscaling directly into your workflow. It connects to API providers (Replicate, fal.ai, Google Gemini) to run state-of-the-art models without leaving Photoshop.

What are the system requirements?​

You need Adobe Photoshop 26.10.0 or later, a Mac or Windows machine, Adobe Creative Cloud Desktop app, and an internet connection. You'll also need your own API keys from at least one provider (Replicate, fal.ai, or Gemini).

Is this a subscription service?​

No, Cubby Image is a one-time payment. You will receive lifetime access to the plugin. It is not a subscription service. To generate images you will need access to API keys from our supported providers and pay for the API usage based on the number of images generated and the pricing of the API provider you choose.

How do I install Cubby Image?​

Simply download the .ccx file and double-click it to install via Adobe's Unified Plugin Installer Agent (UPIA). After installation, restart Photoshop, open the plugin from Plugins → Cubby Image, accept the Terms of Use, and enter your API keys.

For detailed instructions, see the Installation Guide.

What is a .CCX file? Why do you not offer .ZXP or .JSX files?​

A .CCX file is a plugin package format used and recommended by Adobe for installing extensions in Creative Cloud applications. We use .CCX files because they are the standard format for Adobe plugins and ensure a smooth installation process. .ZXP (old plugin format) and .JSX (just scripts) are older formats that do not offer the variety of features needed for Cubby Image.

Can I use Cubby Image with unofficial versions of Photoshop?​

Using unofficial versions may lead to compatibility issues and is not supported. We cannot provide support for such setups.

When installing Adobe Creative Cloud, I see the message: 'Compatible app required. You don't have a compatible app installed for this plugin.' Why?​

This error usually means Creative Cloud cannot detect the correct version of Photoshop.

Solutions:

  1. Make sure you are using at least Photoshop 26.10.0 or later
  2. If you already have the correct version, try reinstalling Photoshop
  3. In some cases, Photoshop may not be properly registered with Creative Cloud, which a reinstall usually fixes

How do I check my Photoshop version?​

  1. Open Photoshop
  2. Go to Photoshop → About Photoshop (Mac) or Help → About Photoshop (Windows)
  3. The version number appears in the splash screen

You need version 26.10.0 or later.

The plugin doesn't appear in my Plugins menu​

Try these steps:

  1. Make sure you installed via the .ccx file (double-click to install)
  2. Restart Photoshop completely
  3. Check Creative Cloud Desktop → Plugins → Manage Plugins to verify it's installed
  4. Ensure the plugin is enabled in that list

API Keys & Billing​

Do I need my own API keys?​

Yes, Cubby Image requires your own API keys from providers like Replicate, fal.ai, or Google Gemini. The plugin does not come with tokens or keys - you provide your own, which gives you full control over costs with no markup.

Are image generations free?​

No. Image generations run through cloud APIs such as Replicate, fal.ai, and Gemini, which involve costs.

To use them, you'll need to:

  1. Create an API key for your preferred provider
  2. Set up billing with that provider
  3. Add the key to Cubby Image

The user manual provides a step-by-step guide to help you set this up.

Typical costs range from $0.001 to $0.10 per image depending on the model.

I'm getting 'Rate limit reached' errors with Gemini API keys. Why?​

You likely haven't set up billing for your Gemini account. The free tier has very limited access and doesn't support image generation. Set up billing in your Google Cloud console or use Google's $300 free credit offer (90 days). Check Gemini's rate limits documentation for current tiers.

Which AI models are available?​

Available models include Nano Banana (Google), Flux Kontext Pro and Max (Black Forest Labs), Seedream 4.0 (ByteDance), QWEN Image Edit (Alibaba), Reve Edit, Topaz Labs upscaling, Real-ESRGAN, Clarity, Crisp, Crystal upscalers, and multiple background removal models. The exact models depend on your API provider.

For a complete list, see the APIs & Models reference.

My API key isn't working. Why?​

This usually comes down to formatting or saving issues.

Solutions:

  1. After pasting your key, make sure you clicked Save
  2. Double-check that you pasted the correct key
  3. Verify there are no extra spaces before or after the key
  4. Click the Show (eye) button to confirm the key is entered correctly
  5. Try deleting and re-entering the key

Other causes:

  • Your API provider account may not have billing set up
  • The key may have been revoked in your provider's dashboard
  • Your account may be out of credits or have payment issues

How much do API calls cost?​

Costs vary by provider and model. Typical ranges:

  • Nano Banana: $0.001 - $0.01 per image
  • Seedream 4.0: $0.02 - $0.05 per image
  • Flux Kontext: $0.02 - $0.08 per image
  • Topaz Upscale: $0.01 - $0.10 depending on scale

Check your API provider's pricing page for exact current rates.

Can I use multiple API providers?​

Yes! You can set up API keys for all three providers (Replicate, fal.ai, Gemini) and switch between them when generating images.

Generation Issues​

My generation failed. Why?​

There can be many reasons why a generation fails.

First steps:

  1. Try again - temporary network issues can cause failures
  2. Check your internet connection
  3. Verify your API key is still valid

Common causes:

  • Content filters: Some models have stricter content filters. Try switching to a different model or provider
  • Network issues: Temporary connectivity problems
  • API provider outage: Check the provider's status page
  • Prompt issues: Adjust your prompt or context images

If failures persist, try:

  • Using a different model
  • Simplifying your prompt
  • Removing or changing context images
  • Switching to a different API provider

I'm getting slow results. Why?​

Processing speed depends on the model and API provider.

Expected times:

  • Nano Banana: Usually fast (10-30 seconds)
  • Seedream 4.0: Usually takes longer (30-90 seconds)
  • Upscaling: Can take 1-5 minutes depending on scale factor

Some delay is normal, especially for high-quality models. Be patient for best results.

If unusually slow:

  • Check the provider's status page for issues
  • Try a different provider
  • Reduce the number of variants
  • Try during off-peak hours

The quality of my results isn't good. Why?​

Different models require different prompting styles and techniques.

Solutions:

  1. Read best practices for the specific model you're using
  2. Be more detailed and specific in your prompts
    • ❌ "a dog"
    • ✅ "a golden retriever puppy sitting in green grass, natural sunlight, professional photography, shallow depth of field"
  3. Try different models to find which works best for your use case
  4. Use context images to guide the results
  5. Generate multiple variants to explore different interpretations

General prompting tips:

  • Describe the subject in detail
  • Specify the style (photorealistic, illustration, cartoon, etc.)
  • Include lighting details (natural light, studio lighting, golden hour, etc.)
  • Mention composition (close-up, wide angle, from above, etc.)
  • Add quality modifiers (high detail, 4k, professional, etc.)

Images appear on the wrong layer​

Generated images appear as new layers in your document. They don't replace or modify existing layers.

To organize:

  1. Use layer groups to organize generated images
  2. Rename layers for clarity
  3. Delete unwanted generations

This non-destructive workflow means your original work is always preserved.

Upscaling Issues​

Upscaling is very slow​

This is normal behavior. Upscaling, especially at higher scale factors, requires significant processing:

  • 2x upscale: 30-60 seconds
  • 4x upscale: 1-3 minutes
  • 6x upscale: 3-5 minutes or more

The quality improvement is worth the wait for most professional use cases.

Upscaling ran out of memory​

Large upscaling operations can require substantial memory:

Solutions:

  1. Try a lower scale factor (2x instead of 4x)
  2. Close other applications to free up memory
  3. Upscale in sections if working with very large images
  4. Start with a smaller source image

Upscaling didn't improve quality​

If upscaling doesn't help:

  • The source image may be too low quality or heavily compressed
  • Try cleaning up noise/artifacts before upscaling
  • Some AI-generated images may already be at optimal quality
  • Try different upscaling settings in Advanced Settings

Context Images & Selections​

My selection isn't being used​

Make sure:

  1. You have an active selection before generating
  2. The selection is on the current document
  3. You haven't accidentally clicked Select All which overrides it

Context images aren't having an effect​

Tips for effective context images:

  1. Use high-quality images as context
  2. Order matters - try reordering your context images
  3. Some models respond better to context than others
  4. Make sure your prompt references the context (e.g., "in the style of the provided image")

How many context images can I add?​

While there's no hard limit in the interface, consider:

  • More images = longer processing time
  • Most effective: 1-3 context images
  • Too many can confuse the model

Settings & Configuration​

Can I view my saved API keys?​

No. For security reasons, saved API keys cannot be displayed in Cubby Image.

If you need to verify or change a key:

  1. Click Delete to remove the current key
  2. Paste a new key
  3. Click Show (eye icon) before saving to verify
  4. Click Save

My settings aren't saving​

Solutions:

  1. Make sure Photoshop has write permissions
  2. Restart Photoshop
  3. Check you're running a supported version (26.10.0+)
  4. Temporarily disable security software that might block Photoshop
  5. Try running Photoshop as administrator (Windows)

Do I need to accept Terms again after updates?​

Sometimes, yes. If there are significant changes to the Terms of Use, you may need to review and accept them again after a plugin update.

Performance & Workflow​

Can I generate while Photoshop is in the background?​

Yes, but:

  • Keep Photoshop running (don't quit)
  • The plugin panel should remain open
  • You can work in other applications while it generates

How do I organize generated layers?​

Best practices:

  1. Use layer groups to organize by prompt or session
  2. Rename layers with descriptive names
  3. Delete unwanted generations immediately
  4. Use layer comps to save different variations

Can I use generated images commercially?​

This depends on:

  • The model's license (check model documentation)
  • Your API provider's terms
  • Your account type with the provider

Always review the specific model's license and your provider's terms before using generated images commercially.

Error Messages​

"API request failed"​

Causes:

  • Network connectivity issues
  • Invalid API key
  • API provider outage
  • Insufficient credits/billing issues

Solutions:

  1. Check your internet connection
  2. Verify your API key is correct
  3. Check your provider's status page
  4. Verify your billing is set up

"Model not available"​

Causes:

  • The model may be temporarily unavailable
  • The model may have been deprecated
  • Your API provider may not support this model

Solutions:

  1. Try a different model
  2. Switch to a different API provider
  3. Check the APIs & Models page for current availability

"Selection required"​

Some operations require an active selection:

  1. Use any Photoshop selection tool
  2. Create a selection on your canvas
  3. Or click Select All in Cubby Image

Still Need Help?​

If your question isn't answered here:

  1. Review the generating images guide
  2. Review the upscaling guide
  3. Check the APIs & Models reference
  4. Check your API provider's documentation
  5. Contact support at support@cubbytools.com

When contacting support, please include:

  • Photoshop version
  • Cubby Image version
  • Operating system
  • API provider being used
  • Error messages (if any)
  • Steps to reproduce the issue