TEST #3873 RESULTS
07/07/2026 - 9:35 PM
ACCELERATOR
MODEL
29.0
tokens/s
generation
11.03
sec
time to first token
143
tokens/s
prompt
72
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core i7-6700K CPU @ 4.00GHz (skylake)
RAM
60.5GB
OS
Linux
Kernel Release
6.14.0-37-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #37-Ubuntu SMP PREEMPT_DYNAMIC Fri Nov 14 22:10:32 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
139
tokens/s
32.6
tokens/s
7.41
sec
pp4096+tg256
106
tokens/s
23.4
tokens/s
38.74
sec
pp2048+tg256
124
tokens/s
27.1
tokens/s
16.60
sec
pp2048+tg768
126
tokens/s
27.3
tokens/s
16.32
sec
pp1024+tg1024
128
tokens/s
28.7
tokens/s
8.01
sec
pp1280+tg3072
140
tokens/s
24.3
tokens/s
9.21
sec
pp384+tg1152
156
tokens/s
31.2
tokens/s
2.49
sec
pp64+tg1024
189
tokens/s
33.0
tokens/s
367
ms
pp16+tg1536
178
tokens/s
33.1
tokens/s
117
ms