PingYi

PingYi contributors·Qingshihuan.PingYi

Privacy-first screenshot OCR and translation desktop tool

PingYi captures a selected screen region, recognizes its text, and translates it in a compact result card. It supports fully local OCR and translation, local LLM services such as llama.cpp and Ollama, and optional Google Cloud or Baidu providers using credentials supplied by the user. The default local workflow keeps screenshots and text on the device and does not retain a history.

winget install --id Qingshihuan.PingYi --exact --source winget

Latest 0.3.0·August 12, 2026

Release Notes
  • Added English and Simplified Chinese interfaces with automatic system-language selection.
  • Added Google Cloud Vision OCR and Google Cloud Translation using a user-provided API key.
  • Kept local PaddleOCR, offline translation, llama.cpp, and Ollama workflows available.

Installer type: inno

x64—5266DB16C9C502278C58959B7F4C314873BCF10153ED3A6A4AA27D378036289C

Details

Homepage
https://github.com/qingshihuan/pingyi
License
MIT
Publisher
PingYi contributors
Support
https://github.com/qingshihuan/pingyi/issues
Moniker
pingyi

Tags

capturegoogle-cloud-visionllama-cpplocal-llmocrofflineollamascreen-translatorscreenshottext-recognitiontranslation