Conversational chat
Render chatbot output progressively while the model is still generating.
AI Tools · Const
Streaming Responses
Streaming exposes parts of an AI response as soon as they are generated, so a UI can begin rendering earlier than with a blocking response. It improves perceived responsiveness for conversational interfaces.
How it works
Streaming sends portions of a model response as soon as they are available instead of waiting for the complete response. In the AI SDK, streamText produces an asynchronous text stream that consumers can iterate over incrementally.
Syntax
streamText({ model, prompt })Returns
An incremental stream of response text consumed as chunks become available.
Key points
- Streaming delivers response fragments incrementally, allowing the interface to show output before generation completes.
- Streaming improves perceived responsiveness because users can see the beginning of an answer sooner.
- A non-streaming interface cannot render partial model output while the response is still being generated.
- Choose streaming or blocking behavior according to the interface requirements and their respective trade-offs.
Common mistakes
- A blocking response delays all visible output until generation finishes, so use streaming when early rendering matters.
Example: Stream text incrementally
import { streamText } from 'ai';
import { openai } from '@ai-sdk/openai';
const result = streamText({
model: openai('gpt-4o-mini'),
prompt: 'Explain streaming in one sentence.',
});
for await (const chunk of result.textStream) {
process.stdout.write(chunk);
}Streams generated text chunks to the terminal as they arrive.
When to use it
Long responses
Show the first useful words quickly in a long-form generation workflow.
Responsive user interfaces
Reduce perceived waiting time in an interactive AI interface.
Related
streamTexttextStreamuseChatuseCompletion