A Windows local screenshot OCR, error translation and screenshot Q&A tool with automatic v2rayN proxy or direct connection
- mobile-v0.1.0-alpha.1
- Published at
- 22d ago
- 1
- Stars
- 9
- Download
- Synced at
- 9d ago
Latest release
mobile-v0.1.0-alpha.1Sep 27, 2026, 07:11 PMPre-releaseDownloads slow? Switch the source anytime above
README
Project Overview
ScreenLingo is a Windows desktop tool focused on local screenshot OCR, error translation and screenshot Q&A. Select an English code error on screen and release the mouse to get local recognition and a translation, making English errors easy to understand quickly. Built with Electron, its OCR runs locally and normal translation only sends the recognized text; to keep asking about a screenshot, you can configure an OpenAI-compatible vision model.
Select an English error and see the explanation in Chinese the moment you release the mouse.
Key Features
- Screenshot translation:
Alt+Shift+Sto select and translate, with local OCR plus online translation - Clipboard translation:
Alt+Shift+Ttranslates already-copied text directly - Screenshot Q&A: keeps asking about the error cause or fix direction; requires a vision model
- Adaptive networking: detects v2rayN proxy or direct connection automatically, or set it manually
- Tray resident: stays in the system tray after closing, ready whenever you need it
Target Audience
- Developers who frequently deal with English code errors
- Learners and office workers who need quick on-screen translation
- Local developers working in v2rayN proxy environments
Quick Start
- Download the latest version from Releases, unzip and run
截译 ScreenLingo.exe - Choose a translation service in Settings on first use
- Press
Alt+Shift+Sin any window and drag to select an English error - Release the mouse and wait for the local OCR and translation result
Core Advantages
- Local OCR: recognition runs on your machine with a bundled English model, balancing speed and privacy
- One-shot translation: one shortcut from selection to result
- Proxy adaptive: works whether v2rayN is on or off, with no manual switching
System Requirements
| Item | Requirement |
|---|---|
| OS | Windows |
| Screenshot Q&A | An OpenAI-compatible vision model required |
| Default translation | Network required |
Notes
- Screenshots are sent to the vision model only when you actively use "screenshot Q&A"; normal translation sends text only
- API keys are encrypted with Windows secure storage and never fall back to plain text
- The unsigned build may show a SmartScreen prompt; only download from Releases
Release history
Sort