- Added support for the OpenRouter API (allows you to use over 300 models, including models from OpenAI, Google, xAI, Meta)
- Added option to set the model to use for Gemini
- Fixed and improved DeepL implementation (it's hard for me to verify if stuff works for DeepL, because I can't use their API, so if stuff is still broken, please tell me)
Linguji is only compiled for Windows for now, but it should also work on Linux and macOS (except video support), so feel free to compile it yourself.
Builds for Linux and macOS will follow.