Returning tokens as they are generated instead of waiting for completion.
1 min readupdated 2026-07-04
/ quick answer
Streaming (SSE or chunked HTTP) lets the UI render partial output within ~200ms, dropping perceived latency dramatically even when total generation time is unchanged. Returning tokens as they are generated instead of waiting for completion.
Returning tokens as they are generated instead of waiting for completion. Streaming (SSE or chunked HTTP) lets the UI render partial output within ~200ms, dropping perceived latency dramatically even when total generation time is unchanged. In practice: ChatGPT's typewriter effect is server-sent event streaming. This dictionary node is part of the Onexial knowledge graph and links to related concepts, workflows and tools below.
Definition
Streaming (SSE or chunked HTTP) lets the UI render partial output within ~200ms, dropping perceived latency dramatically even when total generation time is unchanged.
Example
ChatGPT's typewriter effect is server-sent event streaming.
/ frequently asked
What is Streaming?
Streaming (SSE or chunked HTTP) lets the UI render partial output within ~200ms, dropping perceived latency dramatically even when total generation time is unchanged.
What is an example of Streaming?
ChatGPT's typewriter effect is server-sent event streaming.
Why does Streaming matter for AI and automation?
Returning tokens as they are generated instead of waiting for completion. It connects to the workflows, prompts and tool stacks linked on this page, so you can move from definition to execution without leaving Onexial.