Streaming Text

useCompletion — hook streaming AI responses in real time.

Try a prompt:

Ready

Click a prompt above to see streaming in action

The code:

import { useCompletion } from "@inferface/hooks";

const { completion, complete, isLoading, abort } =
  useCompletion({ api: "/api/complete" });

// Fire a completion
await complete("Explain async/await");

// Stream renders automatically via `completion`

Zero dependencies. @inferface/hooks handles SSE parsing, abort, error recovery, and state management. Works with OpenAI, Anthropic, or any SSE-compatible endpoint.