Realtime

Run updates#

Run updates provide status, metadata, and tags as a run changes. The React hook is useRealtimeRun.

import { useRealtimeRun } from "@trigger.dev/react-hooks";
 
const { run, error } = useRealtimeRun(runId, {
  accessToken: publicAccessToken,
});

Output: the component receives the current run object and re-renders as the run changes.

Streams#

Streams carry continuous data, such as AI tokens, file chunks, or progress. Define a stream in the task, then consume it with useRealtimeStream.

const stream = streams.define("answer");
await stream.pipeText(llmResponse);

Output: the frontend can read data as the task produces it.

Run updates and streaming#

You can use both together: show a progress indicator with run updates and stream the response beside it.

Authentication#

Realtime hooks and functions require authentication. Use the upstream Realtime authentication guide for the current token and scope setup.

Next steps#