🚀 New Model​

Nano Banana Pro​

We've added Nano Banana Pro, Google's newest multimodal image creation model powered by Gemini 3. This model offers extreme definition and enhanced capabilities compared to the original Nano Banana.

Nano Banana Pro is available through:

• Replicate
• fal.ai
• Google Gemini API

Experience next-level image generation with improved detail, better prompt understanding, and stunning output quality.

🚀 New Feature​

Crystal Upscaler on fal.ai​

The ClarityAI Crystal upscaler is now available through fal.ai, giving you another provider option for premium upscaling with crystal-clear results.

🐛 Bug Fixes​

Context Tile UI Fix​

Fixed context tile UI rendering issues across Mac & Windows, as well as across different Photoshop versions. The interface now displays consistently regardless of your platform or PS version.

Why Figma?​

Figma has become the go-to tool for collaborative design work. Whether you're creating interfaces, designing marketing materials, or prototyping products, having AI-powered image generation right inside Figma eliminates app-switching and keeps your creative flow uninterrupted.

All Your Favorite Features, Now in Figma​

Cubby Image for Figma launches with feature parity to our Photoshop plugin, including:

🎨 Image Generation​

Generate stunning images from text prompts using state-of-the-art AI models:

• Gemini 2.5 (Nano Banana) - Google's latest image generation model
• Seedream 4.0 - ByteDance's powerful AI with 1K, 2K, and 4K quality options
• Flux Kontext Pro & Max - Black Forest Labs' advanced generation
• Reve Edit, Remix & Create - Versatile editing and creation tools

🔍 Image Upscaling​

Enhance your images with multiple upscaling models:

• Topaz Labs - Professional-grade upscaling
• Recraft Crisp - Sharp detail enhancement
• ClarityAI Clarity & Crystal - Intelligent enhancement
• Real-ESRGAN - Proven upscaling technology

✂️ Background Removal​

Remove backgrounds with precision using advanced AI models:

• Recraft, lucataco, BiRefNet v2, Bria RMBG 2.0, 851 Labs

🎯 Reference Images​

Use your Figma frames and selections as references for AI generation, maintaining consistency across your designs.

⚙️ Full Control​

• Custom aspect ratios for all models
• Optional prompt enhancement toggle
• Cancel operations mid-generation
• Direct frame creation for generated images

Your Keys, Your Control​

Just like the Photoshop version, Cubby Image for Figma uses your own API keys from providers like Replicate, fal.ai, and Google Gemini. No subscriptions, no per-image fees from us - you pay providers directly with full transparency and control.

Getting Started​

1. Install from Figma Community - Visit the Cubby Image plugin page and click "Open in..."
2. Add Your License - Log in with your Gumroad license (or purchase a new one)
3. Connect API Keys - Add your API keys from Replicate, fal.ai, or Google Gemini
4. Start Creating - Generate, upscale, and edit images directly in your Figma files

Special Launch Offer​

To celebrate the Figma launch, we're offering a Bundle deal with both Photoshop and Figma plugins at 25% off. Early adopters can lock in lifetime access before we transition to a different pricing model.

Get Cubby Image for Figma →

What's Next?​

We're committed to evolving both plugins with new features and models. Your feedback has been instrumental in shaping Cubby Image, and we can't wait to see what you create with it in Figma.

Have questions or feedback? Reach out to us - we'd love to hear from you!

Ready to supercharge your Figma workflow? Try Cubby Image for Figma today and experience AI image generation without leaving your design tool.

🚀 New Functionality​

Background Removal​

Remove backgrounds from your images with intelligent masking logic for precise cutouts. Choose from multiple advanced models:

• Recraft - Fast and accurate background removal
• lucataco - High-quality cutouts with fine detail
• BiRefNet v2 - Advanced boundary refinement
• Bria RMBG 2.0 - Professional-grade background removal
• 851 Labs - Optimized for complex subjects

New Upscaling Models​

We've added four new upscaling models to enhance your images with different approaches:

• Recraft Crisp - Sharp, detailed upscaling for graphics and illustrations
• ClarityAI Clarity - Balanced enhancement for general use
• ClarityAI Crystal - Enhanced detail preservation
• Real-ESRGAN - Proven technology for photorealistic upscaling

New Image Editing Models​

Expand your creative possibilities with powerful new editing capabilities:

• Reve Edit - Precise image editing with AI guidance
• Reve Remix - Transform and reimagine existing images
• Reve Create - Generate variations with creative control

Licensing System​

Log into Cubby Image directly with your Gumroad license for seamless activation. No more manual key entry - just sign in and start creating.

✨ User Experience Improvements​

Reorder Context Images​

We've completely redesigned our tile reordering algorithm, making drag-and-drop reordering of context images smooth and reliable.

Aspect Ratio Flexibility​

Aspect ratio controls are now available for all supported models, giving you complete control over your image dimensions.

Quality Controls for Seedream 4.0​

Choose your desired output quality when using Seedream 4.0:

• 1K - Fast generation for quick iterations
• 2K - Balanced quality and speed
• 4K - Maximum quality for final outputs

Create in New Document​

Generate images directly into new Photoshop documents for a cleaner workflow. Perfect for starting fresh projects or organizing multiple generations.

Cancel Operations​

Need to stop a generation mid-process? The new cancel button lets you interrupt running operations instantly.

Disable Prompt Enhancement​

Take full control with the option to disable automatic prompt enhancement. Great for when you want your prompts processed exactly as written.

Enhanced Button Layout​

We've refined the action button positioning and visual hierarchy for a more intuitive interface.

Larger Prompt Input​

The prompt text area is now significantly larger, making it easier to write and edit complex prompts.

Get the Update​

Existing users can download v0.3.0 from their Gumroad purchase page. New to Cubby Image? Get started today with a lifetime license.

As always, we'd love to hear your feedback! Reach out through our contact page or join the conversation in our community.

Why Nano Banana Understands You Better​

Unlike traditional text-to-image models that rely on keyword matching, Nano Banana has deep language understanding. This means you can—and should—write prompts like you're describing a scene to another person, not just listing tags.

Instead of this:

cat, coffee shop, morning, cozy, warm lighting

Try this:

A photorealistic wide-angle shot of a tabby cat lounging on a windowsill 
in a cozy coffee shop. Morning sunlight streams through the window, 
creating warm, golden highlights on the cat's fur. The background shows 
blurred espresso machines and chalkboard menus.

The difference? Context and narrative. Nano Banana thrives on descriptive paragraphs.

The Six Essential Prompting Patterns​

1. Photorealistic Scenes: Think Like a Photographer​

For realistic images, use actual photography terminology. Mention camera angles, lens types, and lighting setups.

Template:

A [shot type] of [subject], [action or expression], set in [environment]. 
The scene is illuminated by [lighting description], creating a [mood] 
atmosphere. Captured with a [camera/lens details], emphasizing 
[key textures and details].

Real Example:

A close-up portrait of an elderly Japanese ceramicist carefully shaping 
a clay bowl on a pottery wheel. Natural window light from the left creates 
dramatic shadows, emphasizing the texture of the clay and the craftsman's 
weathered hands. Shot with an 85mm lens at f/2.8, shallow depth of field.

2. Product Photography: Studio-Quality Results​

Perfect for e-commerce or marketing materials without hiring a photographer.

Template:

A high-resolution, studio-lit product photograph of [product description] 
on a [background surface]. The lighting is [lighting setup] to highlight 
[specific feature]. The camera angle is [angle type]. Ultra-realistic, 
with sharp focus on [key detail].

Pro tip: Specify the lighting setup (three-point lighting, softbox, ring light) for more controlled results.

3. Illustrations and Stickers: Be Explicit About Style​

For icons, stickers, or design assets, clarity is everything.

Template:

A [style] sticker of a [subject], featuring [key characteristics] and 
a [color palette]. The design should have [line style] and [shading style]. 
The background must be transparent.

Example:

A kawaii-style sticker of a happy red panda holding a coffee cup, 
featuring big sparkly eyes and a cute smile. Use a pastel color palette 
with pink and orange tones. The design should have bold black outlines 
and flat cel shading. The background must be transparent.

4. Text in Images: Nano Banana's Superpower​

One of Nano Banana's standout features is accurate text rendering—something most AI models struggle with.

Template:

Create a [image type] with the text "[exact text]" in a [font style]. 
The design should be [style description], with a [color scheme].

Example:

Create a modern, minimalist logo for a coffee shop called 'The Daily Grind' 
with the text in a bold, geometric sans-serif font. The design should feature 
a stylized coffee cup integrated into the 'D', with a warm brown and cream 
color scheme.

5. Minimalist Compositions: Perfect for Presentations​

Great for creating backgrounds where text will be overlaid.

Template:

A minimalist composition featuring [subject] positioned in the 
[placement] of the frame. The background is a vast, empty [color] 
canvas, creating significant negative space. Soft, subtle lighting.

6. Sequential Art: For Visual Storytelling​

Template:

A single comic book panel in a [art style] style. In the foreground, 
[character description and action]. In the background, [setting details]. 
The panel has a dialogue box with the text "[Text]". The lighting creates 
a [mood] mood.

Advanced Techniques​

Iterative Refinement​

Don't expect perfection on the first try. Generate multiple variations and refine your prompts based on the results. If the first generation is close but needs adjustments, modify your prompt to be more specific about what you want to change.

Step-by-Step Instructions for Complex Scenes​

Break complex prompts into sequential steps:

First, create a background of a serene, misty forest at dawn. 
Then, in the foreground, add a moss-covered ancient stone altar. 
Finally, place a single, glowing sword on top of the altar.

Semantic Negative Prompts​

Instead of saying what you don't want, describe what you do want positively:

❌ "a street with no cars" ✅ "an empty, deserted street with parked bicycles along the sidewalk"

Control the Camera​

Use cinematic language to control composition:

• "wide-angle shot" for expansive scenes
• "macro shot" for extreme close-ups
• "low-angle perspective" to make subjects look imposing
• "bird's eye view" for overhead shots

Working with Context Images​

Nano Banana can use reference images alongside your text prompts. This is perfect for:

Style Transfer:

Transform this photograph of my living room into the artistic style 
of Van Gogh's Starry Night. Preserve the room's layout and furniture 
placement, but render it with bold, swirling brushstrokes and a vibrant 
blue and yellow palette.

Adding Elements:

Using the provided image of my product, place it on a modern marble 
countertop in a bright, minimalist kitchen. Ensure the product's features 
and branding remain completely unchanged.

Inpainting:

Using the provided image, change only the wall color to a sage green. 
Keep everything else exactly the same, preserving the original lighting 
and composition.

Making This Easier with Cubby Image​

If you're working in Photoshop, you can use Nano Banana directly without switching between apps. That's exactly why we built Cubby Image—to integrate AI image generation seamlessly into your Photoshop workflow.

With Cubby Image, you can:

• Access Nano Banana (and other models) right from your Photoshop panels
• Use your active selections as masks automatically
• Pull context images directly from your layers
• Generate multiple variations side-by-side without leaving your document

The key is having your tools work with your creative flow, not against it. When you're deep in a project, the last thing you want is to export files, switch to a browser, generate images, then import them back. Cubby Image keeps everything in one place.

Best Practices Checklist​

✅ Be hyper-specific - More detail = more control ✅ Provide context - Explain the image's purpose ✅ Use narrative descriptions - Not just keyword lists ✅ Iterate and refine - Generate multiple variations with adjusted prompts ✅ Think photographically - Use real camera/lighting terms ✅ Break down complexity - Step-by-step for complex scenes

Common Pitfalls to Avoid​

❌ Listing disconnected keywords ❌ Being vague about style or mood ❌ Forgetting to specify aspect ratio ❌ Not mentioning lighting for realistic scenes ❌ Using negative language ("no cars") instead of positive ("empty street")

Real-World Applications​

For Designers​

• Product mockups for client presentations
• Quick concept visualization
• Marketing material backgrounds
• Logo variations and iterations

For Content Creators​

• Social media graphics
• Blog post hero images
• Thumbnail designs
• Branded illustrations

For Photographers​

• Mood board creation
• Pre-visualization of complex shots
• Background replacement concepts
• Style reference generation

The Bottom Line​

Nano Banana's strength lies in its ability to understand natural language and detailed descriptions. The more specific and descriptive your prompts, the better your results will be.

Start with descriptive, narrative prompts. Use photography and artistic terminology. Generate multiple variations and refine your prompts based on results. Most importantly, experiment—every prompt teaches you something new about how to communicate with AI.

What has your experience been with AI image generation? Found any prompting techniques that work particularly well? I'd love to hear about them.

Try These Techniques in Your Workflow​

Ready to use Nano Banana in Photoshop? We built Cubby Image specifically to bring these AI models directly into your creative environment. No more switching between apps, exporting files, or losing context.

With Cubby Image, you get:

• Nano Banana (plus FLUX, Qwen, and more) right in Photoshop
• Automatic context from your selections and layers
• Quick iteration by generating variations without re-exporting
• Side-by-side comparison of variations
• One-click upscaling with AI upscalers

The prompting techniques you learned here work even better when you can iterate quickly without breaking your creative flow.

Get Cubby Image → | Read the Docs

Questions about prompting or AI image generation? Feel free to reach out—I'm always happy to help troubleshoot or share what I've learned.

What is Nano Banana?​

Nano Banana is an AI model that uses Google's Gemini 2.5 Flash technology to generate high-quality images from text prompts. It's known for:

• ⚡️ Speed: Generates images in seconds
• 🎯 Accuracy: Understands complex prompts well
• 📐 Flexibility: Works with various aspect ratios
• 💰 Cost-effective: Affordable API pricing

Note on Image Quality: Nano Banana generates images at a limited resolution. However, Cubby Image has an integrated AI upscaler, so you can always upscale your results to higher resolutions once you're happy with the generated image!

Getting Started​

Prerequisites​

1. Install Cubby Image
2. Get an API key from any of these providers:
   • Replicate (model: google/nano-banana)
   • fal.ai (model: fal-ai/gemini-nano-banana)
   • Gemini API (direct access)

Basic Usage​

1. Open Cubby Image in Photoshop
2. Select Nano Banana from the model dropdown
3. Enter your prompt
4. Click Generate

Writing Effective Prompts​

Nano Banana understands natural language well. Here are some tips:

Structure Your Prompt​

[Subject] + [Setting] + [Style] + [Details]

Example:

A mystical forest at dawn, 
magical atmosphere with floating lights, 
digital painting style, 
soft purple and blue tones

Use Descriptive Language​

❌ "castle" ✅ "medieval stone castle on a hilltop, surrounded by mist, dramatic lighting"

Specify Art Style​

• photorealistic
• digital art
• oil painting
• watercolor
• 3D render
• anime style

Advanced Techniques​

Controlling Composition​

Use positional terms:

• "in the foreground"
• "centered composition"
• "viewed from above"
• "wide angle shot"

Lighting and Mood​

• golden hour lighting
• dramatic shadows
• soft ambient light
• neon lights
• volumetric fog

Quality Enhancers​

Add these to improve results:

• highly detailed
• 8k resolution
• sharp focus
• professional photography

Real-World Examples​

Product Photography​

High-end wireless headphones on a wooden desk,
minimalist setup, natural window lighting,
product photography, shallow depth of field,
professional commercial style

Landscape Art​

Vast desert landscape with sand dunes,
sunset with orange and pink sky,
single traveler silhouette,
cinematic composition, ultra wide angle

Character Design​

Steampunk inventor character,
Victorian era clothing with brass goggles,
workshop background with gadgets,
detailed illustration, concept art style

Tips and Tricks​

1. Iterate and Refine​

Don't expect perfection on the first try. Generate multiple variations and refine your prompt.

2. Use Seeds for Consistency​

If you find a good result, note the seed number to generate similar images.

3. Combine with Photoshop Tools​

• Use generated images as a base layer
• Add manual touch-ups with brushes
• Combine multiple generations
• Apply Photoshop filters and adjustments

4. Aspect Ratios​

Nano Banana works well with:

• Square (1024x1024) - Social media
• Portrait (768x1024) - Vertical designs
• Landscape (1024x768) - Wide scenes

Common Issues and Solutions​

Image Not Matching Prompt​

• Make your prompt more specific
• Break complex ideas into simpler descriptions
• Try adding "professional" or "high quality"

Unwanted Elements​

• Use negative prompts if available
• Specify what you DON'T want
• Regenerate with adjusted prompt

Consistency Issues​

• Use the same seed value
• Keep core prompt elements the same
• Only change specific details

Cost Optimization​

Nano Banana is cost-effective, but here are ways to optimize:

1. Draft locally: Plan your composition in Photoshop first
2. Batch similar prompts: Generate variations together
3. Use lower steps: For drafts, fewer steps = faster & cheaper
4. Refine manually: Use Photoshop for small adjustments instead of regenerating

Comparison with Other Models​

Conclusion​

Nano Banana is an excellent all-around model for AI image generation in Photoshop. Its combination of speed, quality, and cost-effectiveness makes it perfect for:

• Quick concept exploration
• Social media content
• Iteration and experimentation
• Learning AI generation

Start experimenting with Nano Banana today and discover what you can create!

Try Cubby Image: Download now

Have questions? Email us at support@cubbytools.com

✨ What's new?​

🖼️ QWEN Image Edit model​

QWEN's brand-new image model delivers high-fidelity generation and editing, now on par with Nano Banana and Seedream 4.0.

🔗 fal.ai API support​

Working on fal.ai? You can now access all currently available models through their API.

🎯 Select All button​

No more frustration selecting the whole canvas — just click the new "Select All" button below the Generate button.

🔄 Reuse prompt & settings​

Iterating just got easier. With a generative layer selected, you can reuse prompts and settings (Selection, Prompt and Context Images) with one click. (Only available per session, once the document is closed they cannot be retrieved)

🎨 UI improvements​

Action buttons refined, plus the "remove reference image" button is now easier to find (including on Windows).

📕 User Manual​

We created a user manual for anyone that needs some help setting up Cubby Image on their machine. We will continually fill it with more information and frequently asked questions.

🙏 A big thank you!​

We're grateful for your support and feedback — it's been amazing to hear how much you enjoy Cubby Image. We'll keep building, improving, and adding new models and features.

We're already working on the next update. Currently we're planning to add advanced settings like aspect ratio, image size and general model specific settings as well as new upscaler models. Stay tuned!

Cheers,
Your Cubby Team
Ludger & Kenny

What to Expect​

On this blog, we'll share:

• 📚 Tutorials: Step-by-step guides for using different AI models
• 💡 Tips & Tricks: Pro tips to get better results
• 🆕 Updates: New features and model additions
• 🎨 Showcases: Amazing work from our community
• 📰 News: Industry updates and AI developments

Stay Connected​

• Website: cubbytools.com
• Documentation: docs.cubbytools.com
• Email: support@cubbytools.com

Happy creating! 🎨✨

After weeks of intensive development, we're excited to bring you a tool that seamlessly integrates cutting-edge AI models into your creative workflow.

🚀 What is Cubby Image?​

Cubby Image is a Photoshop plugin that brings powerful AI image generation and upscaling capabilities right into your workspace. No need to switch between applications — generate, edit, and upscale images without ever leaving Photoshop.

✨ Key Features​

🎨 Multiple AI Models​

Access leading AI models including:

• Nano Banana - Fast, high-quality image generation
• Seedream 4.0 - Advanced generation with excellent prompt adherence
• And more models being added regularly!

🖼️ Context-Aware Generation​

• Use selections to control where AI generates content
• Reference images for style matching and inspiration
• Smart prompting for precise results

📈 AI Upscaling​

Enhance your images with AI-powered upscaling to increase resolution while maintaining quality.

⚡ Seamless Workflow​

• Generate directly on new layers
• Preserve your original work
• Iterate quickly with intuitive controls

🔑 Multi-Provider Support​

Connect your own API keys from:

• Replicate
• Google Gemini

🎯 Who is it for?​

Whether you're a digital artist, photographer, designer, or creative professional, Cubby Image helps you:

• Quickly generate concepts and variations
• Fill in missing content with AI
• Upscale images for print or high-res displays
• Experiment with AI-powered creativity

🛠️ Getting Started​

Getting started is simple:

1. Purchase Cubby Image on Gumroad
2. Follow our installation guide
3. Set up your API keys
4. Start creating!

🙏 Thank You!​

We're incredibly excited to share Cubby Image with you. This is just the beginning — we have many more features and improvements planned.

If you have any questions, feedback, or suggestions, please don't hesitate to reach out. We'd love to hear from you!

Happy creating!

The Cubby Team
Ludger & Kenny