Real-time React SDK

React hooks for interacting with the Speechmatics Real-time transcription API.

This package wraps the @speechmatics/real-time-client package for use in React projects.

Set Up

1. Create an account on the Speechmatics On-Demand Portal here.
2. Navigate to Manage > API Keys page in the Speechmatics On-Demand Portal.
3. Enter a name for your API key and store your API key somewhere safe.

Installlation

npm i @speechmatics/real-time-client-react

info

For React Native, make sure to install the event-target-polyfill package, or any other polyfill for the EventTarget class

Usage

Below is an example of usage in the browser.

1. Use the RealtimeTranscriptionProvider at a high level of your component tree:

   import { RealtimeTranscriptionProvider } from "@speechmatics/real-time-client-react";
   import { PCMAudioRecorderProvider } from '@speechmatics/browser-audio-input-react';
   
   function RootLayout({children}) {
     /*
       Two context providers wrapping the app:
         - One for the Speechmatics Real time transcription client
         - One for capturing PCM microphone audio in the browser
           see https://www.npmjs.com/package/@speechmatics/browser-audio-input-react
     */
     return <RealtimeTranscriptionProvider appId="your-app-id">
       <PCMAudioRecorderProvider workletScriptURL="/js/pcm-audio-worklet.min.js">
         {children}
       </PCMAudioRecorderProvider>
     </RealtimeTranscriptionProvider>
   }
   

   Note thatRealtimeTranscriptionProvider is a client component, like any other context provider. In NextJS, it's best to put this either in a root layout, or inside another client component. For frameworks like Remix which don't use React Server Components, it should work anywhere.

2. Inside a component below the RealtimeTranscriptionProvider:

   import {
     type RealtimeTranscriptionConfig,
     useRealtimeTranscription,
   } from '@speechmatics/real-time-client-react';
   
   import {
     usePCMAudioListener,
     usePCMAudioRecorder,
   } from '@speechmatics/browser-audio-input-react';
   
   // We recommend 16_000Hz sample rate for speech audio.
   // Anything higher will be downsampled server-side
   const RECORDING_SAMPLE_RATE = 16_000;
   
   function MyComponent() {
     const { startTranscription, stopTranscription, sendAudio, socketState } =
       useRealtimeTranscription();
   
     const { isRecording, startRecording, stopRecording } = usePCMAudioRecorder();
   
     // Send audio to Speechmatics when captured
     usePCMAudioListener(sendAudio);
   
     const startSession = useCallback(
       async (config: RealtimeTranscriptionConfig) => {
         // getJWT can fetch an ephemeral key based on your setup
         // See our NextJS example: https://github.com/speechmatics/speechmatics-js-sdk/blob/main/examples/nextjs/src/app/actions.ts
         const jwt = await getJWT('rt');
   
         // Start a Speechmatics session, then start recording to stream the audio
         await startTranscription(jwt, config);
         await startRecording({ sampleRate: RECORDING_SAMPLE_RATE });
       },
       [startTranscription, startRecording],
     );
   }
   

Source on GitHub

The source code for this package can be found on GitHub.