TEST #733 RESULTS
04/17/2025 - 8:55 AM
ACCELERATOR
MODEL
27.6
tokens/s
generation
1.68
sec
time to first token
791
tokens/s
prompt
235
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
12th Gen Intel Core i7-12700H (alderlake)
RAM
31.6GB
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
871
tokens/s
28.0
tokens/s
1.21
sec
pp4096+tg256
721
tokens/s
25.2
tokens/s
5.72
sec
pp2048+tg256
848
tokens/s
27.6
tokens/s
2.45
sec
pp2048+tg768
844
tokens/s
27.2
tokens/s
2.46
sec
pp1024+tg1024
914
tokens/s
27.8
tokens/s
1.15
sec
pp1280+tg3072
890
tokens/s
26.6
tokens/s
1.47
sec
pp384+tg1152
960
tokens/s
28.5
tokens/s
434
ms
pp64+tg1024
801
tokens/s
28.6
tokens/s
115
ms
pp16+tg1536
270
tokens/s
28.4
tokens/s
93
ms