TEST #3617 RESULTS
05/26/2026 - 1:51 PM
ACCELERATOR
MODEL
93.3
tokens/s
generation
431
ms
time to first token
2950
tokens/s
prompt
861
LocalScore
HOW YOU STACK UP
Explore All ResultsMeta Llama 3.1 8B Instruct - Q4_K - Medium
SYSTEM
CPU
AMD Ryzen 7 5700X 8-Core Processor (znver3)
RAM
30.3GB
OS
Linux
Kernel Release
7.0.0-15-generic
Architecture
x86_64
Version
Cosmopolitan 3.9.7 MODE=x86_64; #15-Ubuntu SMP PREEMPT_DYNAMIC Wed Apr 22 16:06:43 UTC 2026
RUNTIME
Name
llamafile
Version
0.9.2
Commit Hash
a30b324
DETAILED RESULTS
TEST NAME
PROMPT
GENERATION
TTFT
pp1024+tg16
2719
tokens/s
96.3
tokens/s
387
ms
pp4096+tg256
2921
tokens/s
84.2
tokens/s
1.41
sec
pp2048+tg256
3403
tokens/s
91.8
tokens/s
613
ms
pp2048+tg768
3368
tokens/s
91.0
tokens/s
619
ms
pp1024+tg1024
3657
tokens/s
94.7
tokens/s
291
ms
pp1280+tg3072
3522
tokens/s
89.7
tokens/s
374
ms
pp384+tg1152
3707
tokens/s
96.4
tokens/s
114
ms
pp64+tg1024
2491
tokens/s
98.1
tokens/s
35
ms
pp16+tg1536
760
tokens/s
97.3
tokens/s
31
ms