TEST #1289 RESULTS
08/02/2025 - 1:22 PM
ACCELERATOR
MODEL
74.9
tokens/s
generation
721
ms
time to first token
1828
tokens/s
prompt
575
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core i5-4590S CPU @ 3.00GHz (haswell)
RAM
23.9GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.3
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2186
tokens/s
79.5
tokens/s
481
ms
pp4096+tg256
1633
tokens/s
51.5
tokens/s
2.53
sec
pp2048+tg256
1976
tokens/s
68.8
tokens/s
1.05
sec
pp2048+tg768
1972
tokens/s
65.8
tokens/s
1.05
sec
pp1024+tg1024
2172
tokens/s
77.0
tokens/s
483
ms
pp1280+tg3072
2084
tokens/s
62.1
tokens/s
627
ms
pp384+tg1152
2262
tokens/s
86.2
tokens/s
180
ms
pp64+tg1024
1561
tokens/s
93.5
tokens/s
51
ms
pp16+tg1536
610
tokens/s
89.6
tokens/s
35
ms