A desktop tool that lets you use various AI models to translate, summarize, and explain anything on your screen through screenshots. No copy-paste needed.
When you encounter these situations, give it a try:
- Foreign web pages, software interfaces, or PDF papers you can't read: Take a screenshot with a hotkey and get instant translation.
- Text you can't copy (like text in images or scanned PDFs): Take a screenshot and accurately extract and translate the text.
- Complex charts or screenshots with lots of information: Take a screenshot and let AI summarize the key points, explain the principles, or extract structured text.
The tool has three built-in "pipelines" for processing images, suitable for different scenarios:
- Quick Translation Pipeline: The easiest option - AI directly returns the translated full text after you take a screenshot.
- Precise Extraction Pipeline: First recognizes all text in the image with high precision, then translates or processes it. Perfect for complex content with formatting or formulas.
- Free Q&A Pipeline: Converts the screenshot content into a structured JSON description of the image, then the language model understands the JSON description. You can command AI to do more, like: "Summarize the core points of this image in Chinese", "List the operation steps in the image", "Explain this flowchart".
- Download: Download the latest installation package from release.
- Install: Run the installer just like any other software.
- Configure (Optional): If you want to use your own AI API, you can fill them in the settings.
- Use: Press the default
Alt+A(configurable) to take a screenshot, select an area, and the result will appear in a floating window next to it.
Essentially, it's a local client that handles screenshots, displays results, and "stitches" together your chosen AI services. The core is the "PIPELINE" design:
User screenshot → Select processing method → Tool calls the corresponding AI service → Result display
You can freely configure which AI service to use at each step (cloud API or local model).
- Is it free? The tool itself is free. But if you use paid third-party AI APIs (like GPT-4), you'll incur the corresponding costs.
- Is my screenshot data secure? If you configure cloud AI services, images will be sent to the corresponding service provider. If you use local models exclusively, your data stays entirely on your computer.
- Why is the installation package a bit large? Because it includes some necessary runtime environments and basic local processing models to ensure some functionality works even without an internet connection.
If you want to build from source:
# Clone the repository
git clone https://github.com/aorucshiea/VisionBridge.git
cd VisionBridge
# Install dependencies
npm install
# Build the application
npm run build
# Run the application
npm run devI created this tool originally to solve the trouble of reading foreign language materials myself. The code is open source. If you have similar needs or want to improve it, you're welcome to contribute code or ideas.