TEST #3591 RESULTS
05/21/2026 - 6:09 PM
ACCELERATOR
MODEL
79.7
tokens/s
generation
377
ms
time to first token
3363
tokens/s
prompt
892
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 9 5950X 16-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
0.9.3
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
4152
tokens/s
83.2
tokens/s
259
ms
pp4096+tg256
3365
tokens/s
77.4
tokens/s
1.23
sec
pp2048+tg256
3856
tokens/s
79.7
tokens/s
544
ms
pp2048+tg768
3604
tokens/s
76.8
tokens/s
582
ms
pp1024+tg1024
4094
tokens/s
81.1
tokens/s
263
ms
pp1280+tg3072
3946
tokens/s
79.7
tokens/s
337
ms
pp384+tg1152
4187
tokens/s
77.2
tokens/s
105
ms
pp64+tg1024
2346
tokens/s
80.2
tokens/s
39
ms
pp16+tg1536
719
tokens/s
82.0
tokens/s
34
ms