TEST #1765 RESULTS
11/09/2025 - 1:43 AM
ACCELERATOR
MODEL
40.7
tokens/s
generation
795
ms
time to first token
1702
tokens/s
prompt
443
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 5 5600 6-Core Processor (znver3)
RAM
31.9GB
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
1976
tokens/s
41.7
tokens/s
542
ms
pp4096+tg256
1510
tokens/s
38.8
tokens/s
2.74
sec
pp2048+tg256
1800
tokens/s
40.0
tokens/s
1.16
sec
pp2048+tg768
1786
tokens/s
40.7
tokens/s
1.17
sec
pp1024+tg1024
1988
tokens/s
41.1
tokens/s
539
ms
pp1280+tg3072
1926
tokens/s
39.6
tokens/s
689
ms
pp384+tg1152
2206
tokens/s
40.9
tokens/s
198
ms
pp64+tg1024
1612
tokens/s
41.6
tokens/s
63
ms
pp16+tg1536
516
tokens/s
41.5
tokens/s
54
ms