TEST #1582 RESULTS
10/05/2025 - 8:27 AM
ACCELERATOR
MODEL
24.0
tokens/s
generation
2.60
sec
time to first token
543
tokens/s
prompt
171
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core i7-5775C CPU @ 3.30GHz (broadwell)
RAM
15.6GB
OS
Linux
Kernel Release
6.8.0-85-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #85-Ubuntu SMP PREEMPT_DYNAMIC Thu Sep 18 15:26:59 UTC 2025
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
598
tokens/s
27.9
tokens/s
1.75
sec
pp4096+tg256
453
tokens/s
15.0
tokens/s
9.12
sec
pp2048+tg256
540
tokens/s
21.9
tokens/s
3.84
sec
pp2048+tg768
540
tokens/s
20.8
tokens/s
3.84
sec
pp1024+tg1024
596
tokens/s
24.6
tokens/s
1.76
sec
pp1280+tg3072
576
tokens/s
19.1
tokens/s
2.26
sec
pp384+tg1152
633
tokens/s
27.7
tokens/s
639
ms
pp64+tg1024
616
tokens/s
30.4
tokens/s
133
ms
pp16+tg1536
335
tokens/s
28.9
tokens/s
75
ms