Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Time-to-first-token (TTFT) on Llama 8B Instruct 64K input length 3.1
Loading...
7.2
TTFT (s)
TriangleMix + FlexPrefill
7.04
8.12
9.2
10.28
Jul 29, 2025
TTFT (s)
Updated 1mo ago
Evaluation Results
Method
Method
Links
TTFT (s)
TriangleMix + FlexPrefill
2025.07
7.2
TriangleMix + MInference
2025.07
7.7
FlexPrefill
2025.07
7.7
TriangleMix + XAttention
2025.07
7.7
TriangleMix
2025.07
8.6
DuoAttention
2025.07
8.7
XAttention
2025.07
9.4
MInference
2025.07
10.1
Dense
2025.07
11.2
Feedback
Search any
task
Search any
task