Hey… Can we please get Text Modality or Support Voice Cloning and SSML for 3.1 Live? Those voices really aren’t going to cut it and trying to hack the transcription to use Cartersia.ai just isn’t working.
\ I believe the current future for Gemini is being put through some last tunings and will be geared toward a particular niche. There are many models out there but you seem to be looking for a model that is specialized in the matter you specified. Have you tried looking for any models that are geared for that purpose?
Gemini Live 2.0 exp and 2.5 then 4.9 release supported text modality, were faster and fit the niche, it’s with the push to STS and full cascade that isn’t working. 3.1 Live is slow, the voices can’t be personalized and it’s expensive. Feels like we are going backwards not forward with Gemini live…