Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Time-to-first-token (TTFT) on Llama-3.1-8B-Instruct 80K input length
Loading...
9.2
TTFT (s)
TriangleMix + FlexPrefill
8.932
10.741
12.55
14.359
Jul 29, 2025
TTFT (s)
Updated 1mo ago
Evaluation Results
Method
Method
Links
TTFT (s)
TriangleMix + FlexPrefill
2025.07
9.2
TriangleMix + MInference
2025.07
9.5
FlexPrefill
2025.07
9.9
TriangleMix + XAttention
2025.07
10
DuoAttention
2025.07
11.7
TriangleMix
2025.07
11.7
MInference
2025.07
12.3
XAttention
2025.07
12.4
Dense
2025.07
15.9
Feedback
Search any
task
Search any
task