AI and LLMs · Middle · Question 36 of 75
What does time to first token (TTFT) measure?
- 1How quickly the first output token arrives
- 2How long the whole answer takes to finish
- 3How long the user takes to read the answer
Show the answer
Correct answer: How quickly the first output token arrives
Explanation
TTFT drives perceived speed in chat. Long prompts, large models and cold caches increase it.