Discussions

Ask a Question

Preventing Avatar's Default Responses in Voice Chat Mode?

In voice chat mode, I'm using streamingAvatar.speak({ text: responseFromClaude, task_type: TaskType.REPEAT }) to have the avatar repeat back a response from Claude AI. However, the avatar still responds from its own default knowledge base before repeating Claude's message, leading to two responses.

ANSWERED

Intermittent Reconnection Issue with Interactive Avatar Session

I’m encountering an intermittent issue when opening a session with an interactive avatar using the SDK. This doesn't happen every time, but occasionally, after starting the session, the system begins trying to reconnect, as shown in the attached log.

ANSWERED

How do I fix this error and why is it occuring?

Locally, when I debug my project, the microphone works properly and all good, when I deploy it on IIS and try to open it, the microphone sign on the browser tab is shown, but the microphone itself is not.

ANSWERED

Access Token Usage

Can anyone initialize streaming avatar in frontend from my access token? What about the access token generated from enterprise account where I got from this https://help.heygen.com/en/articles/9182113-interactive-avatar-101-your-ultimate-guide that When you use the Enterprise API Token in concert with the Streaming API, you can create up to 100 concurrent sessions, with unlimited monthly usage and session length.

ANSWERED

How to increase session_duration_limit

Hi,

ANSWERED

Streaming API endpoint for avatar.list changed the response.

Hello,

Can I have two avatars on screen at the same time having a dialog?

And can I do this via API.

Getting 500 errors from Trial Token key

Error starting avatar session:
Error {name: "APIError", status: 500, responseText: "{"code":500,"message":"something wrong in server"}"}

Avatar speaking random things on it's own in the new SDK.

Hi, I started using the new sdk today. First of all, the heygen team has done a great job with reducing the latency of streaming.task api. However, it tends to speak random sentence after every 15 seconds, even if streaming.task api is in action?