TEST #3167 RESULTS
03/23/2026 - 6:46 PM
ACCELERATOR
MODEL
59.9
tokens/s
generation
1.69
sec
time to first token
887
tokens/s
prompt
316
LocalScore
HOW YOU STACK UP
Explore All ResultsLlama 3.2 1B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 9 7945HX with Radeon Graphics (znver4)
RAM
60.5GB
OS
Linux
Kernel Release
6.8.0-106-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #106-Ubuntu SMP PREEMPT_DYNAMIC Fri Mar 6 07:58:08 UTC 2026
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
951
tokens/s
63.1
tokens/s
1.09
sec
pp4096+tg256
654
tokens/s
53.0
tokens/s
6.28
sec
pp2048+tg256
842
tokens/s
59.0
tokens/s
2.45
sec
pp2048+tg768
843
tokens/s
58.2
tokens/s
2.45
sec
pp1024+tg1024
970
tokens/s
60.4
tokens/s
1.07
sec
pp1280+tg3072
944
tokens/s
56.7
tokens/s
1.37
sec
pp384+tg1152
1019
tokens/s
62.2
tokens/s
393
ms
pp64+tg1024
1030
tokens/s
63.8
tokens/s
77
ms
pp16+tg1536
730
tokens/s
62.8
tokens/s
37
ms