TEST #1839 RESULTS
11/21/2025 - 4:33 PM
ACCELERATOR
MODEL
118
tokens/s
generation
185
ms
time to first token
7455
tokens/s
prompt
1680
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Xeon Gold 5220R CPU @ 2.20GHz (cascadelake)
RAM
63.6GB
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
9713
tokens/s
128
tokens/s
113
ms
pp4096+tg256
6321
tokens/s
76.5
tokens/s
660
ms
pp2048+tg256
8117
tokens/s
129
tokens/s
260
ms
pp2048+tg768
8195
tokens/s
83.7
tokens/s
263
ms
pp1024+tg1024
9673
tokens/s
129
tokens/s
114
ms
pp1280+tg3072
9018
tokens/s
129
tokens/s
150
ms
pp384+tg1152
11352
tokens/s
127
tokens/s
42
ms
pp64+tg1024
4199
tokens/s
129
tokens/s
24
ms
pp16+tg1536
509
tokens/s
128
tokens/s
39
ms