TEST #2953 RESULTS
03/02/2026 - 12:21 PM
ACCELERATOR
MODEL
77.3
tokens/s
generation
270
ms
time to first token
4924
tokens/s
prompt
1122
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen Threadripper PRO 5945WX 12-Cores (znver3)
RAM
127.9GB
OS
Windows
Kernel Release
10.0
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64
RUNTIME
Name
llamafile
Version
0.9.3
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
6058
tokens/s
76.4
tokens/s
182
ms
pp4096+tg256
4462
tokens/s
73.2
tokens/s
933
ms
pp2048+tg256
5457
tokens/s
77.5
tokens/s
388
ms
pp2048+tg768
5495
tokens/s
75.9
tokens/s
387
ms
pp1024+tg1024
6110
tokens/s
76.9
tokens/s
180
ms
pp1280+tg3072
6061
tokens/s
75.6
tokens/s
224
ms
pp384+tg1152
6574
tokens/s
79.6
tokens/s
71
ms
pp64+tg1024
3129
tokens/s
80.4
tokens/s
32
ms
pp16+tg1536
973
tokens/s
80.0
tokens/s
28
ms