TEST #986 RESULTS
06/03/2025 - 4:09 AM
ACCELERATOR
MODEL
26.0
tokens/s
generation
1.65
sec
time to first token
813
tokens/s
prompt
234
LocalScore
HOW YOU STACK UP
Explore All ResultsDeepSeek R1 Distill Llama 8B - Q4_K - Medium
SYSTEM
CPU
Intel Core i5-4590 CPU @ 3.30GHz (haswell)
RAM
31.9GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
941
tokens/s
26.8
tokens/s
1.13
sec
pp4096+tg256
738
tokens/s
23.8
tokens/s
5.60
sec
pp2048+tg256
865
tokens/s
25.8
tokens/s
2.41
sec
pp2048+tg768
847
tokens/s
25.4
tokens/s
2.46
sec
pp1024+tg1024
934
tokens/s
26.4
tokens/s
1.14
sec
pp1280+tg3072
897
tokens/s
25.1
tokens/s
1.47
sec
pp384+tg1152
1011
tokens/s
26.8
tokens/s
418
ms
pp64+tg1024
821
tokens/s
27.0
tokens/s
116
ms
pp16+tg1536
259
tokens/s
26.9
tokens/s
101
ms