TEST #374 RESULTS
04/06/2025 - 2:03 PM
ACCELERATOR
MODEL
20.7
tokens/s
generation
2.58
sec
time to first token
567
tokens/s
prompt
166
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
Intel Core i7-10750H CPU @ 2.60GHz (skylake)
RAM
31.8GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
621
tokens/s
25.2
tokens/s
1.69
sec
pp4096+tg256
440
tokens/s
7.2
tokens/s
9.44
sec
pp2048+tg256
534
tokens/s
13.6
tokens/s
3.91
sec
pp2048+tg768
560
tokens/s
13.0
tokens/s
3.73
sec
pp1024+tg1024
645
tokens/s
19.2
tokens/s
1.63
sec
pp1280+tg3072
628
tokens/s
11.4
tokens/s
2.08
sec
pp384+tg1152
699
tokens/s
27.8
tokens/s
570
ms
pp64+tg1024
697
tokens/s
38.2
tokens/s
103
ms
pp16+tg1536
281
tokens/s
30.7
tokens/s
69
ms