Verified85% confidenceFactTime unknown
TPS measurement is typically divided into two phases: Time to First Token (TTFT) and the generation rate (throughput) of subsequent tokens.
5
Sources
85%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedTPOT (Time Per Output Token) refers to the average time taken to generate each subsequent token, and total response time is roughly equal to TTFT + TPOT × number of output tokens.79% similarUnverifiedPerformance timing in streaming output scenarios requires measuring three dimensions separately: time to first byte (TTFT), final completion time, and intermediate chunk throughput rate78% similarUnverifiedtps (tokens per second) is a core metric for measuring model inference speed: below 30 tps you will clearly feel the wait, 60-100 tps can satisfy most dialogue and coding needs, and above 100 tps is quite smooth.78% similarUnverifiedResponse speed involves time to first token (TTFT) and overall throughput.76% similarUnverifiedWhen measuring LLM time to first token (TTFT), the timer should be stopped when the first token is seen, and then the remaining stream should continue to be consumed, rather than waiting until the entire response ends.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/18528API
curl https://kongchang.com/api/v1/knowledge/claims/18528MCP
get_claim(id=18528)