By skyforbes 
By using the combination of Gemini3+ChatGPT, I have achieved full AI implementation from technology selection to coding.
I utilized the libreTranslate open-source translation engine and SoVit.
This enables automatic translation of the input text into the target language during chatting and simultaneous interpretation during voice calls.
However, due to the open-source solutions I used and the insufficient performance of my computer, the experience of the voice call function is not good.
My next step is to optimize the voice function and then develop the simultaneous interpretation function for video calls.