TEST #3199 RESULTS
03/26/2026 - 4:00 AM
ACCELERATOR
MODEL
7.5
tokens/s
generation
2.64
sec
time to first token
506
tokens/s
prompt
113
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 7840HS with Radeon 780M Graphics (znver4)
RAM
29.7GB
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
633
tokens/s
8.3
tokens/s
1.74
sec
pp4096+tg256
472
tokens/s
2.3
tokens/s
9.17
sec
pp2048+tg256
550
tokens/s
5.5
tokens/s
3.93
sec
pp2048+tg768
564
tokens/s
4.9
tokens/s
3.84
sec
pp1024+tg1024
627
tokens/s
7.2
tokens/s
1.75
sec
pp1280+tg3072
595
tokens/s
4.0
tokens/s
2.29
sec
pp384+tg1152
641
tokens/s
10.2
tokens/s
664
ms
pp64+tg1024
253
tokens/s
14.0
tokens/s
294
ms
pp16+tg1536
216
tokens/s
11.4
tokens/s
110
ms