Handling Mixed-Language Named Entities in Russian TTS Output

Hello,

I am researching Russian text-to-speech systems, particularly for cases where English named entities (like “KFC”, “Google”, etc.) are embedded in Russian sentences. Although I have not used the vosk-tts model directly yet, I have experience with other TTS models trained with VITS. In my experiments, those models could not generate natural-sounding audio for such mixed-language input—English words were pronounced using Russian phonetic rules, which affects intelligibility and naturalness.

I am interested in whether vosk-tts can handle these cases better, or if there are any plans to support more natural code-switching/mixed-language pronunciation in the future. Proper handling of named entities is very important for many real-world applications.

Do you have any suggestions or workarounds? Is this feature on the roadmap for vosk-tts? Any advice or pointers would be greatly appreciated.

Thank you for your work on this project!

Best regards,
afun