Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Time To First Token (TTFT) on LLM Prefix-based Inference
Loading...
51
TTFT (ms)
FlashRT
-29.24
512.38
1,054
1,595.62
Jun 18, 2026
TTFT (ms)
Updated 1mo ago
Evaluation Results
Method
Method
Links
TTFT (ms)
FlashRT
prefix_length=2048, ca...
2026.06
51
FlashRT
prefix_length=4096, ca...
2026.06
53
FlashRT
prefix_length=8192, ca...
2026.06
54
vLLM
prefix_length=4096, ca...
2026.06
76
vLLM
prefix_length=8192, ca...
2026.06
120
vLLM
prefix_length=2048, ca...
2026.06
143
FlashRT
prefix_length=2048, ca...
2026.06
200
FlashRT
prefix_length=4096, ca...
2026.06
365
vLLM
prefix_length=2048, ca...
2026.06
519
FlashRT
prefix_length=8192, ca...
2026.06
723
vLLM
prefix_length=4096, ca...
2026.06
1,026
vLLM
prefix_length=8192, ca...
2026.06
2,057
Feedback
Search any
task
Search any
task