Queries on RADTTS Multilingual Support, SSML Testing, and MagpieTTS

Hello,

I have a few queries regarding RADTTS and MagpieTTS. For RADTTS, does it currently support multilingual dubbing? If not, is there a planned roadmap or recommended workaround for implementing this? Additionally, I noticed that SSML tags appear to require an SSL certificate — is there any way to test SSML locally without SSL? For MagpieTTS, I would like to know if it’s possible to request preview access for evaluation purposes. Also, are there any plans to support SSML, text substitution, or emotion mixing features in MagpieTTS? Thank you for your guidance.

RADTTS model was archived. Magpie-TTS is the current recommended model.
MagpieTTS - the multilingual model is publicly available. SSML is not supported, but the model can pickup the right way to pronounce the sentence. In the future, Pace and pause will be introduced.
for zeroshot - you can apply for access here : Log in | NVIDIA Developer and add the NVIDIA contact person to confirm the request.

Emotions are supported with basic 5 emotions for the model with OOTB voices. The Zero-shot will mimic the audio sample pitch and tone.