xxnuo/MTranServer
MTranServer
MTranServer is a self-hosted translation server that runs entirely offline using locally stored models, processing language conversions without any internet connection or GPU hardware. It functions as a translation API emulator, mimicking the endpoints of popular translation services so that existing client software can connect without requiring configuration changes.
The server is designed for private, local deployment, packaged as a containerized backend for consistent installation across different environments. It supports multiple API protocols, making it compatible with browser translation plugins like Immersive Translate and DeepL, as well as IDE extensions for translating code comments and strings. The server can be configured entirely through command-line flags at startup, and it includes system tray integration for running as a background desktop service with a simple web UI for management.
The project covers the full workflow of offline machine translation, from serving translations via local models to integrating with browser extensions and development tools. It provides a complete self-hosted alternative to cloud-based translation APIs, keeping all data local and eliminating dependency on external services.
Features
- Translation Servers - Runs a private translation service on your own hardware to keep data local and avoid cloud API costs or internet dependency.
- Local Model Serving - Serves translations using locally stored models, eliminating the need for internet connectivity or GPU hardware.
- Translation Servers - Runs local models to convert text between languages without internet access.
- Offline Translation Engines - Translates text between languages using local models without any internet connection, ensuring privacy and availability.
- Translation Engines - Processes language conversions using locally stored models without requiring GPU hardware.
- Translation API Integrations - Exposes multiple API endpoints compatible with popular browser translation extensions like Immersive Translate and DeepL.
- API Emulators - Mimics popular translation service APIs so existing client software connects without configuration changes.
- Translation Protocol - Exposes multiple API endpoints compatible with various browser translation extensions and IDE plugins.
- AI Translation Extensions - Connects browser translation extensions to a local backend for seamless, private in-page translation.
- Translation API Endpoints - Mimics popular translation service endpoints for compatibility with existing clients.
- Containerized Backends - Provides a containerized deployment for consistent installation and management of translation models across environments.
- Containerized Deployments - Packages the server and models into containers for consistent installation across different host machines.
