Deployment & Inference · Level 3 of 5
Latency
The elapsed time for a request or operation to complete.
Tail latency can matter more than the average for user experience.
Example
A response arrives 120 milliseconds after a request starts.
Listen to the definition and example
Audio transcript
Latency. The elapsed time for a request or operation to complete. Tail latency can matter more than the average for user experience. For example: A response arrives 120 milliseconds after a request starts.
Explore this concept
Why it matters
This helps you balance prediction quality with memory, latency, and throughput.
Related concepts
Quick recall question
Try answering before looking back at the definition.