Discussions

Hello,

Hi, i am testing AI Real-Time Conversation with avatar.
Is it possible to programmatically append to the conversation some information
For examples user is playing some game , and i want to enrich context of the Avatar, so he/she would know what's happening in user's current game session

We are currently researching and testing. I had a few questions:

I'm using the endpoint:

HeyGen’s API appears to support SSML-style voice tags for enhanced voice control, but from my testing and available documentation, it looks like only a limited subset of SSML elements is fully supported. Could you please confirm which SSML tags and attributes are officially supported by the HeyGen API?

While generating videos using the HeyGen module on Make.com, I noticed that there's an option to select a voice, but no option to choose different accents. Does this mean the module doesn't support voice accents?

First, can the streaming avatar be integrated with a Google page?

I have a talking photo avatar that i have added motion and sound effects to through the API. Now when i use this avatar to generate a video with a voice script, only the avatar motion gets applied but not the sound effect. Is there a way to include the sound effect aswell when generating the video?

I’ve been working with Heygen lately and I’m running into a small issue I hope the community can help with. Whenever the avatar is speaking, it tends to get interrupted or cut off before it finishes the sentence, when anyone is talking during the speech of the Avatar.
Does anyone know how to program Heygen (via API or otherwise) to ensure it completes its full speech without being interrupted?
Thanks in advance for any insights!

Okay, so I think it's best.
I explained exactly what I need.
I want to create a automated system and part of that automated system is connecting a GPT to a content generator that can create consistent charicters and scripts through API.
I need a service that has the ability for me to generate a script with my GPT/open AI, to send it through API , and have it generate content for my brand that follows the actions and the scripts..... if the movements are are limited, that's okay, but as long as the character can talk, and take some basic actions then that can work.