TL;DR: Transform your clipboard content with AI using a single hotkey. Translate to multiple languages simultaneously, generate presentation images, or apply any custom AI transformation - all without leaving your workflow.
The Problem: Context Switching is Killing Your Productivity
How many times today have you:
- Copied text to translate it
- Opened a browser tab
- Navigated to ChatGPT or Claude
- Pasted the text
- Copied the result
- Switched back to your original app
- Pasted it back
Now multiply that by every translation, every image generation, every AI-assisted task throughout your day. That's a lot of context switching.
Meet cbfilter: Your AI-Powered Clipboard Assistant
cbfilter is a lightweight Windows utility that brings AI transformation directly to your clipboard. Press Win+Alt+V, select a filter, and watch your clipboard content transform instantly.
What makes it powerful?
- 🔄 Four transformation types: Text→Text, Text→Image, Image→Text, Image→Image
- 🤖 Multiple AI providers: OpenAI, Anthropic Claude, Google Gemini, or any OpenAI-compatible API
- ⚡ Zero context switching: Results paste automatically
- 🎨 Custom filters: Create your own AI transformations
- 🌍 Multilingual: 13 language interfaces out of the box
Real-World Example 1: Simultaneous Multi-Language Translation
Imagine you're preparing content for an international developer community like dev.to. You need your announcement in both Chinese and Japanese.
Traditional workflow:
- Translate to Chinese (3-4 steps)
- Copy result somewhere
- Translate to Japanese (another 3-4 steps)
- Format both translations
- Total: ~10-12 manual steps
With cbfilter:
- Copy your English text
- Press
Win+Alt+Von your app - Select "Translate to Chinese and Japanese"
- Done
Configuration
Here's the magic - a simple custom filter that does simultaneous translation:
Configuration screen for simultaneous translation filter
In Action
Watch it translate dev.to community guidelines to Chinese and Japanese in real-time
The entire process took 17 seconds, but the time depends on the model performance. No tab switching, no copy-paste gymnastics.
Real-World Example 2: Generate Presentation Images Instantly
You're writing documentation and need a visual representation. Instead of opening DALL-E or Midjourney in a browser:
With cbfilter:
- Type your image description
- Copy it
- Press
Win+Alt+Von your app - Select "Generate Image"
- Done
Configuration
Configuration for image generation
In Action
From text description to presentation-ready image in seconds
Generated Images
Getting Started
Installation
- Download the latest release from GitHub
- Run the installer - it automatically sets up based on your system language
- Configure your API key on first launch
-
Start transforming with
Win+Alt+V
Initial Setup
The setup wizard makes configuration painless:
Choose your AI provider and the most affordable models are configured automatically
Default models:
- Text or Vision:
gpt-5-nanoorgemini-2.5-flash-lite - Image Generation & Edit:
gpt-image-1-miniorgemini-2.5-flash-image
Creating Custom Filters
The real power is in customization. Each filter needs:
- Title: What appears in your menu
- Input/Output Type: Text or Image
- Model: Which AI to use
- Prompt: Your transformation instructions
Example: Code Reviewer
Review this code for:
1) bugs,
2) performance issues,
3) security vulnerabilities,
4) best practices.
Be concise but specific.
Example: Screenshot to Markdown Table
Extract the table from this image and convert it to a well-formatted Markdown table.
Example: Manga image from text
Generate a promotional image and story as manga, which passionately introduce the content of the given text.
Why Developers Love It
1. API Flexibility
Works with any OpenAI-compatible API:
- OpenAI
- Anthropic
- Google Gemini
- OpenRouter
- LiteLLM
- Ollama
2. Stays Out of Your Way
- Runs from system tray
- Minimal memory footprint
- No UI unless you need it
- Automatic paste after transformation
3. Open Source & Extensible
- MIT License
- Written in modern C++20
- API provider templates for easy expansion
4. Multilingual by Design
Interface available in 13 languages:
English, Japanese, German, Spanish, French, Italian, Korean, Dutch, Portuguese, Russian, Thai, Vietnamese, Chinese
Technical Highlights
- Modern C++ Implementation: Leverages C++20 features for performance
- Windows SDK Integration: Native clipboard monitoring
-
Persistent Configuration: JSON-based config in
%APPDATA% - Hotkey Management: System-wide hotkey registration
- Image Handling: Base64 encoding for vision models
- Error Resilience: Graceful degradation on API failures
Common Use Cases
For Content Creators:
- Translate blog posts to multiple languages
- Generate featured images from titles
- Summarize long articles
- Extract text from screenshots
- Convert simple hand-drawn illustrations into professional photo-like images
For Developers:
- Review code snippets
- Translate source code to another programming language
- Generate documentation images
- Translate error messages
- Extract data from screenshots (tables, logs, etc.)
- Generate flowchart or class diagram from source code
For Researchers:
- Summarize papers
- Translate research findings
- Extract data from charts
- Generate visualizations from descriptions
For Multilingual Teams:
- Real-time meeting notes translation
- Simultaneous documentation in multiple languages
- Screenshot OCR in any language
- Quick language verification
What's Next?
The project is actively developed and welcomes contributions:
- 🎯 More API provider templates
- 🔧 Community-shared filter libraries
- 📊 Usage analytics and filter performance tracking
Try It Now
- Star the repo: github.com/masaoki/cbfilter
- Download the latest release
- Share your custom filters - open a PR with your creative use cases!
The next time you find yourself manually copying and pasting between AI chat interfaces, remember: there's a faster way. Your clipboard is more powerful than you think.
What's your biggest clipboard pain point? Share in the comments, and let's brainstorm filters to solve it!
Top comments (0)