SpeedExplorer：纯 AI 编写的 Windows 文件管理器，内置本地 OCR 翻译图片查看器
misn-bit/SpeedExplorer
SpeedExplorer 是一款完全由 AI 智能体在人工指导下编写的 Windows 文件管理器，内置图片查看器，可通过 LM Studio 服务器调用本地模型实现 OCR 与翻译，并将识别文本和译文直接叠加在图片上、另存为独立文本文件。
SpeedExplorer is a purely vibe-coded Windows file manager replacement. As in, it’s fully written by AI agents under my guidance.
It's a file manager with a built-in image viewer that has OCR+Translation capabilities using locally run AI in LM Studio through the LM Studio server. The image viewer can show OCR'd text and translation right on top of the image, and it saves the OCR'd text and the translation in a separate text file.
For good quality OCR, i recommend using Gemma-4-31B or better.
It's best to disable the "Enable Thinking" toggle in the server inference tab in LM Studio, unless you prefer the model to potentially think about the OCR for thousands of tokens.
Development
Build the application with:
dotnet build --no-restore
Run the automated tests with:
dotnet test SpeedExplorer.Tests\SpeedExplorer.Tests.csproj --no-restore
Disclaimer & Safety
This software is provided "AS IS", without warranty of any kind.
This project is an experiment in AI-assisted development. The source code was generated by AI models and has not been manually audited line-by-line by a human developer.
- While the application has been tested for stability and safety, unintended behaviors may occur.
- Always backup critical data before using experimental file management tools.
- The author is not liable for any data loss or damages arising from the use of this software.
来源：GitHub · github.com