TEST #3904 RESULTS
07/14/2026 - 4:30 PM
ACCELERATOR
MODEL
5.8
tokens/s
generation
30.10
sec
time to first token
45
tokens/s
prompt
21
LocalScore
HOW YOU STACK UP
Explore All ResultsQwen2.5 14B 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
48
tokens/s
6.1
tokens/s
21.62
sec
pp4096+tg256
41
tokens/s
5.0
tokens/s
100.15
sec
pp2048+tg256
45
tokens/s
5.7
tokens/s
45.21
sec
pp2048+tg768
45
tokens/s
5.6
tokens/s
45.32
sec
pp1024+tg1024
48
tokens/s
5.9
tokens/s
21.61
sec
pp1280+tg3072
48
tokens/s
5.4
tokens/s
27.05
sec
pp384+tg1152
50
tokens/s
6.0
tokens/s
7.89
sec
pp64+tg1024
48
tokens/s
6.2
tokens/s
1.49
sec
pp16+tg1536
36
tokens/s
6.1
tokens/s
609
ms