TEST #3903 RESULTS
07/14/2026 - 3:25 PM
ACCELERATOR
MODEL
10.4
tokens/s
generation
16.20
sec
time to first token
85
tokens/s
prompt
38
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B 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
88
tokens/s
11.0
tokens/s
11.66
sec
pp4096+tg256
76
tokens/s
9.1
tokens/s
54.03
sec
pp2048+tg256
85
tokens/s
10.3
tokens/s
24.28
sec
pp2048+tg768
85
tokens/s
10.2
tokens/s
24.22
sec
pp1024+tg1024
88
tokens/s
10.5
tokens/s
11.73
sec
pp1280+tg3072
89
tokens/s
9.7
tokens/s
14.49
sec
pp384+tg1152
91
tokens/s
10.7
tokens/s
4.31
sec
pp64+tg1024
87
tokens/s
11.1
tokens/s
805
ms
pp16+tg1536
74
tokens/s
11.1
tokens/s
305
ms