Server-sent events
A simple protocol where the server pushes updates to the client over a long-lived HTTP connection.
Think of it like
Like a news ticker — you tune in and the server pushes updates; you don't ask for each one.
Example
A chat API uses SSE so the model's response tokens arrive one-by-one as soon as they're ready.
How it actually works
SSE is HTTP-based (so it works through proxies and firewalls), uses a simple text format (each line is a JSON event), and handles reconnection automatically. It's one-way (server to client), so it's simpler than WebSockets but can't do bidirectional streaming.
For product teams
SSE is the standard for streaming LLM responses — easy to use from browsers and backends.
For engineers
SSE is built into browsers; no library needed. Standard format: `data: {...}\n\n`.
Related
- Streaming — Sending responses as they generate.
Read anything AI without the jargon
Look up any term in plain English, or save terms as you read with the free Chrome extension.
Open DecoderAdd to Chrome