TEST #3200 RESULTS
03/28/2026 - 10:49 AM
ACCELERATOR
MODEL
49.1
tokens/s
generation
381
ms
time to first token
3366
tokens/s
prompt
757
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 7800X3D 8-Core Processor (znver4)
RAM
61.9GB
OS
Linux
Kernel Release
6.18.7-76061807-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #202601231045~1769703228~24.04~cb87b5b SMP PREEMPT_DYNAMIC Thu J
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2948
tokens/s
53.3
tokens/s
366
ms
pp4096+tg256
3288
tokens/s
40.1
tokens/s
1.27
sec
pp2048+tg256
4118
tokens/s
47.2
tokens/s
518
ms
pp2048+tg768
4136
tokens/s
46.2
tokens/s
516
ms
pp1024+tg1024
4691
tokens/s
50.1
tokens/s
237
ms
pp1280+tg3072
4451
tokens/s
44.7
tokens/s
307
ms
pp384+tg1152
4628
tokens/s
52.7
tokens/s
102
ms
pp64+tg1024
1556
tokens/s
54.4
tokens/s
58
ms
pp16+tg1536
479
tokens/s
53.4
tokens/s
52
ms