TEST #3257 RESULTS
04/08/2026 - 8:56 PM
ACCELERATOR
MODEL
33.0
tokens/s
generation
10.41
sec
time to first token
136
tokens/s
prompt
76
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core Ultra 5 135H (alderlake)
RAM
30.8GB
OS
Linux
Kernel Release
6.12.0-160000.26-default
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #1 SMP PREEMPT_DYNAMIC Thu Feb 5 00:00:11 UTC 2026 (f762390)
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
163
tokens/s
40.2
tokens/s
6.30
sec
pp4096+tg256
118
tokens/s
32.3
tokens/s
34.71
sec
pp2048+tg256
131
tokens/s
36.8
tokens/s
15.71
sec
pp2048+tg768
117
tokens/s
33.5
tokens/s
17.48
sec
pp1024+tg1024
142
tokens/s
27.3
tokens/s
7.24
sec
pp1280+tg3072
146
tokens/s
30.0
tokens/s
8.77
sec
pp384+tg1152
142
tokens/s
27.1
tokens/s
2.74
sec
pp64+tg1024
119
tokens/s
40.4
tokens/s
573
ms
pp16+tg1536
145
tokens/s
29.0
tokens/s
133
ms