TEST #3902 RESULTS
07/14/2026 - 2:58 PM
ACCELERATOR
MODEL
53.9
tokens/s
generation
3.01
sec
time to first token
512
tokens/s
prompt
209
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 9 7940HS w/ Radeon 780M Graphics (znver4)
RAM
27.1GB
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
528
tokens/s
58.8
tokens/s
1.96
sec
pp4096+tg256
368
tokens/s
43.1
tokens/s
11.13
sec
pp2048+tg256
467
tokens/s
52.1
tokens/s
4.41
sec
pp2048+tg768
470
tokens/s
50.9
tokens/s
4.37
sec
pp1024+tg1024
528
tokens/s
55.2
tokens/s
1.96
sec
pp1280+tg3072
527
tokens/s
48.9
tokens/s
2.44
sec
pp384+tg1152
590
tokens/s
58.2
tokens/s
667
ms
pp64+tg1024
638
tokens/s
59.3
tokens/s
116
ms
pp16+tg1536
490
tokens/s
58.7
tokens/s
49
ms