TEST #4270 RESULTS
10/11/2026 - 7:53 PM
ACCELERATOR
MODEL
28.8
tokens/s
generation
18.49
sec
time to first token
75
tokens/s
prompt
49
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Xeon CPU E5-2630 v2 @ 2.60GHz (ivybridge)
RAM
255.9GB
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
81
tokens/s
30.9
tokens/s
12.71
sec
pp4096+tg256
66
tokens/s
23.7
tokens/s
61.80
sec
pp2048+tg256
73
tokens/s
27.6
tokens/s
28.13
sec
pp2048+tg768
73
tokens/s
27.1
tokens/s
28.08
sec
pp1024+tg1024
78
tokens/s
29.5
tokens/s
13.13
sec
pp1280+tg3072
77
tokens/s
26.3
tokens/s
16.63
sec
pp384+tg1152
81
tokens/s
31.0
tokens/s
4.76
sec
pp64+tg1024
76
tokens/s
32.0
tokens/s
878
ms
pp16+tg1536
73
tokens/s
31.5
tokens/s
249
ms