Discussions

Ask a Question

Voice to voice conversational AI Interactive Avatar

Hello, the online demo where user can speak with an AI ‘Wayne’ in real-time is amazing!

Sometimes the streaming.task API doesn't return 200 for too long? And therefore the avatar doesn't speak

Hi, I am using streaming avatar in my project and i have noticed at times the streaming.task api doesn't even respond back. It neither fails nor give a 200, just stays on pending. Can you give some potential solutions/suggestions on how to prevent this and avoid this behavior of the API, please.

How to generate Interactive HeyGen video using API

I want to generate a video personalized interactive HeyGen Video with some script to speak and I wnat to download that video.Is there any API to create a video, I tried creating video API but it is throwing an error as unautorized and internal error.

Error Retrieving Access Token

import { NextRequest, NextResponse } from "next/server";

Changing voice doesn't seem to work with the @heygen/streaming-avatar library

According to the source code I just gotta set the voice ID:

a couple of questions

do heygen avatars speak as well? can i use another tts software instead?

creating avatar with knowledge base & ai features

can you have it tie into a more in depth knowledge base as well as Ai? to increase the quality of responses? i would want to use it for answering fitness questions but in my style and philosophy of training?

How to retrieve the list of available avatars based on my plan?

Hi HeyGen team,

Videos Stuck in Processing Queue for Almost 48 Hours – No Progress on Translations

I’m facing a critical issue where my videos have been stuck in the processing queue for almost two days without any progress after initiating translations. Every new video I upload also gets stuck in the queue, and nothing is moving forward. This has completely halted my projects, and I'm unable to generate any videos for my clients.

Avatar Speech

How to make the avatar repeat the text which I provide? avatar. Speak() seems to reply the text provide.