Observed arrival · 2026-11-14
IQ-Stream: replayable streams for LLM answers
A server that relays LLM output from an application backend to browsers, with replay when a connection drops.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- Teams streaming LLM answers to browser users
- Worth noticing
- Its proposed flow uses plain HTTP requests for publishing and the browser's built-in EventSource for receiving and reconnecting.
Field notes
The described path runs from an application backend, through HTTP requests carrying JSON lines, to browser clients receiving Server-Sent Events. IQ-Stream says it keeps each answer in memory only during generation and uses the browser's last received event to resume after a reconnect. The page reports 85,000 events per second per CPU and zero missed events in its reconnect benchmark; the supplied extract does not include the benchmark methodology.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
◎NicheUnusually specific use
One card from the complete issue