TEST #1036 RESULTS
06/12/2025 - 2:15 PM
ACCELERATOR
MODEL
65.1
tokens/s
generation
683
ms
time to first token
1898
tokens/s
prompt
566
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 5700G with Radeon Graphics (znver3)
RAM
30.7GB
OS
Linux
Kernel Release
6.11.0-26-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #26~24.04.1-Ubuntu SMP PREEMPT_DYNAMIC Thu Apr 17 19:20:47 UTC 2
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2253
tokens/s
68.0
tokens/s
469
ms
pp4096+tg256
1797
tokens/s
59.6
tokens/s
2.30
sec
pp2048+tg256
2064
tokens/s
64.5
tokens/s
1.01
sec
pp2048+tg768
2061
tokens/s
63.9
tokens/s
1.01
sec
pp1024+tg1024
2220
tokens/s
66.3
tokens/s
476
ms
pp1280+tg3072
2143
tokens/s
61.8
tokens/s
613
ms
pp384+tg1152
2272
tokens/s
66.8
tokens/s
184
ms
pp64+tg1024
1740
tokens/s
68.0
tokens/s
51
ms
pp16+tg1536
530
tokens/s
67.3
tokens/s
44
ms