Use of this sample app is subject to our Terms of Use.
The Zoom Video SDK for Web enables you to build custom video experiences on a webpage with Zoom's core technology. This demo showcases how to use Video SDK RTMS Streams to receive real-time transcripts from Zoom on your backend server. The server then runs a Tensorflow model for Sentiment Analysis on the transcript and outputs the result to the server console.
There is also a Client-side version of this app that uses Zoom Live Transcriptions and Video Preprocessor.
To get started, clone the repo:
git clone https://github.com/zoom/zoom-rtms-sentiment-sample.git
This app requires Node Version 22.22.0 or below. You can install it to your project environment using this command:
nvm install 22
- Launch the Auth Server by clicking below:
| Render |
|---|
 |
-
Retrieve the Auth URL from the deployed Blueprint and input it into the
.envfile. You can also you an Auth URL of another service of your choice. -
Install the dependencies:
bun install # or npm install -
Run the app:
npm run apporbun app
-
Navigate to http://localhost:5173
-
Click "Join" to join the session
-
As you speak, the server will receive RTMS Webhooks containing the transcripts of your speech. The server will then perform a sentiment detection on the text and output a result to the server console
For the full list of features and event listeners, as well as additional guides, see our Video SDK docs.
If you're looking for help, try Developer Support or our Developer Forum. Priority support is also available with Premier Developer Support plans.
Do not expose your credentials to the client, when using the Video SDK in production please make sure to use a backend service to sign the tokens. Don't store credentials in plain text, as this is a sample app we're using an .env for sake of simplicity.