Supercharge Your Clipboard with AI

masaoki — Sun, 21 Dec 2025 08:11:45 +0000

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+V on 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+V on 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-nano or gemini-2.5-flash-lite
Image Generation & Edit: gpt-image-1-mini or gemini-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!

AI #Productivity #OpenSource #Windows #Developer Tools

Sora 2 video generation on Open WebUI

masaoki — Sat, 11 Oct 2025 16:03:47 +0000

Sora 2 API

Currently Sora 2 Web access is available to limited subscribers, but Sora 2 API has been available to developers. Open WebUI is nice place to try Sora 2.

Open WebUI

Open WebUI is a nice front-end to deploy AI models. Docker lets you have personal chat space with your favorite language models. Look at the related article to have your Open WebUI.

Open WebUI extension

Easily installable Open WebUI Function for your Open WebUI to generate Sora 2 or Sora 2 Pro videos.

Writing a better prompt

OpenAI shows a guide to make an effective prompt for Sora 2.
Currently the initial frame is restricted to the non-human image.

Initial frame

Just with a prompt is enough to generate video, but you can specify the image as an initial frame of the generated video. The extension function automatically resize the image to fit the generated video, you still have to take care of the aspect ratio of the image since the image is automatically cropped to fit.

My trial

We will create a commercial video for a laptop specifically designed for dogs. This world begins with an office scene where dogs are working in a bright office using spreadsheet software. The main character, a poodle, takes out a strangely shaped laptop that is shaped like a paw instead of a square, opens it, and quickly completes his work. The other dogs look surprised and ask questions. At the end, the main character says, "Bowtel inside!".

https://x.com/masaoki/status/1977041145671958967

Remix

Remix is available as just continuing the conversation. Edit your video until you're happy with it.

DEV Community: masaoki

Supercharge Your Clipboard with AI

The Problem: Context Switching is Killing Your Productivity

Meet cbfilter: Your AI-Powered Clipboard Assistant

Real-World Example 1: Simultaneous Multi-Language Translation

Configuration

In Action

Real-World Example 2: Generate Presentation Images Instantly

Configuration

In Action

Generated Images

Getting Started

Installation

Initial Setup

Creating Custom Filters

Why Developers Love It

1. API Flexibility

2. Stays Out of Your Way

3. Open Source & Extensible

4. Multilingual by Design

Technical Highlights

Common Use Cases

What's Next?

Try It Now

AI #Productivity #OpenSource #Windows #Developer Tools

Sora 2 video generation on Open WebUI

Sora 2 API

Open WebUI

Open WebUI extension

Writing a better prompt

Initial frame

My trial

Remix