Skip to content

MAI Translator is a multi-ai-support desktop app, allows user to get translations from multiple resources and manage the translations histories locally.

License

Notifications You must be signed in to change notification settings

terrence-ou/MAI-Translator

Repository files navigation

MAI Translator - a multi-ai supported translator

A multi-ai-supported translator helps getting and managing translations with confidence.

Release app Test LICENSE
react typescript react

A translator for users who pursuit translation quality

The MAI Translator helps to retrieve translation results from popular language model providers such as DeepL, Claude, and OpenAI. It allows users to compare different versions of translations and select the words, sentences, or paragraphs that best fit the context. Additionally, the MAI Translator enables users to store translation results locally, which cian be retrieved at any time as needed. Currently the translation histories supports text-to-speech reader.

Features

  1. Fetching translation results from multiple language model providers;
  2. Manage translation records locally;
  3. Light / Dark theme;
  4. Text-to-Speech-supported text reader;

Updates

  • v-0.4.1 & v-0.4.2 09/05/2024
    • AI model configs updated, the user now can select specific models;
    • Added Apple code sign and notary;
  • v-0.4.0 09/01/2024
    • Upgraded Electron to version 32, the latest stable version;
    • Traditional Chinese support (ZH-HANT) added;
  • v-0.3.0
    • Integrated OpenAI TTS model to text reading. (The reading quality is way better than the Chrome reader)
  • v-0.2.1 & v-0.2.2
    • Tooltips added.
    • Translation history becomes expandable by date
  • v-0.2.0
    • Added a navigation menu.
    • Updated translation record interface, displaying the record on a whole page instead of a modal now.
    • Added OpenAI support.

Preparation

For this BETA version, the user needs to provide DeepL free API, OpenAi, and/or Claude API in the Settings to make the translation work.

Sorry, but...

  1. Limited language support due to the language model provider.
  2. Unexpected translation errors might occur.
  3. The software is still under heavy development; only MacOS ARM version available for now.